Does TSDataset support slicing like dataframe? What I what to do is get a sub ts of a TSDataset object, the only way I found is to transform the TSDataset object to a DataFrame, then use .loc() to get a sub dataframe, and finally transform the dataframe to TSDataset again. It seems a liitle tedious.

一使用autots_informer=AutoTS(InformerModel, 24, 6) autots_informer.fit(ts_train, ts_val)
就提示NotImplementedError: Unknown estimator

我用PaddleTS的LSTnet进行预测。在模型fit那里，pycharm报错[paddlets] [ERROR] ValueError: attr: shape doesn't exist!；jupyter notebook内核直接挂掉。但是在飞桨平台就可以正常运行。请问这是什么原因呢？我用MLP和RNN拟合却没问题。

chinese-calendar 最近更新到了1.8.0，支持了2023年**法定节假日安排。

https://github.com/LKI/chinese-calendar/releases/tag/1.8.0

PR #291

spyder 运行官网例子代码：
import numpy as np
import pandas as pd
from paddlets import TSDataset

x = np.linspace(-np.pi, np.pi, 200)
sinx = np.sin(x) * 4 + np.random.randn(200)
df = pd.DataFrame(
{
'time_col': pd.date_range('2022-01-01', periods=200, freq='1h'),
'value': sinx,
'known_cov_1': sinx + 4,
'known_cov_2': sinx + 5,
'observed_cov': sinx + 8,
'static_cov': [1 for i in range(200)],
}
)

target_dataset = TSDataset.load_from_dataframe(
df, #Also can be path to the CSV file
time_col='time_col',
target_cols='value',
freq='1h'
)

输出：
[2022-12-16 18:45:10,564] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!

我有一个predict_result_list，这个list中所有的数据都是有效的。但是，经过paddlets.datasets.tsdataset.TSDataset.concat方法把他们结合起来之后，这里除了第一条数据是正确的之外，剩下的数据都变成了NAN，没有报任何错误。
下面是我的代码：
`

使用test_data预测

predict_result_list = []
for i in paddletsData_list_X:
predicted_dataset = model.predict(i) # 将归一化后的数据集放进去预测
predict_result_list.append(predicted_dataset)

print(predict_result_list[1])

将列表转化为一个paddlets_dataset

total_minmax_result = paddlets.datasets.tsdataset.TSDataset.concat(predict_result_list,axis=1)
total_minmax_result
`

这里面predict_result_list 里面的所有的数据都是有效的，但是使用paddlets.datasets.tsdataset.TSDataset.concat操作之后，得到的结果里面只有第一条数据是正常的，剩下的数据全部都是NAN，时间索引也是正常且连续的。

今天才报名PaddleTS课程学习，报名成功没法看课程回放，显示没有权限，特别想学习PaddleTS，能不能把回放的权限开启，或者怎有没有其他的途径去学习前面的课程？？

    1、问题询问：多时序数据导入时，无法加入其他static_cov_cols变量，因为TSDataset.load中限制了len(np.unique(df[col])) != 1

2、报错信息：raise_if( col not in df.columns or len(np.unique(df[col])) != 1,"static cov cals data is not in columns or schema is not right!")
3、截图示范：见下图

Originally posted by @pvalue31415936 in #110 (comment)

我的out_chunk_len设置为7，predict方法输出为7，怎么才能获得更多时间的预测

问题：utils.py中119行调用的pd.to_timedelta(train_index.freq))方法似乎不支持以月为频率，当我尝试用paddlets对颗粒度为月的数据进行训练时，TSDataset中的freq自动转成或者，导致出现下面的报错信息。

ts = TSDataset.load_from_dataframe(
data, #Also can be path to the CSV file
time_col='time_col',
target_cols=target_field,
known_cov_cols=data.columns.drop(['time_col', target_field]).tolist(),
)
'''
**沿海煤炭运价指数 国际波罗的海干散货海运总和指数 PMI 矿石 焦炭
time_col
2018-01-01 1.49331 1242.000000 51.0 72.227273 2169.090909
2018-02-01 1.09641 1126.157895 50.0 72.769231 1960.000000
2018-03-01 1.05876 1154.238095 52.0 68.136364 1930.454545
2018-04-01 1.08231 1128.900000 51.0 64.764706 1725.500000
2018-05-01 1.26693 1293.095238 52.0 65.523810 1842.045455
2018-06-01 1.21853 1351.619048 52.0 64.500000 2220.000000
......
'''

from paddlets.automl.autots import AutoTS
from paddlets.models.forecasting import MLPRegressor
from ray.tune import uniform, qrandint, choice
from paddlets.transform import Fill

ts_train_val, ts_test = ts.split(0.8)
ts_train, ts_val = ts_train_val.split(0.8)
autots_model = AutoTS(NBEATSModel, 3, 1)
autots_model.fit(train_tsdataset=ts_train_scaled, valid_tsdataset=ts_val_scaled, n_trials=1)
sp = autots_model.search_space()
predicted = autots_model.predict(ts_test_scaled)

报错信息如下：
Failure # 1 (occurred at 2022-11-21_14-41-47)
�[36mray::ImplicitFunc.train()�[39m (pid=9724, ip=127.0.0.1, repr=run_trial)
File "python\ray_raylet.pyx", line 859, in ray._raylet.execute_task
File "python\ray_raylet.pyx", line 863, in ray._raylet.execute_task
File "python\ray_raylet.pyx", line 810, in ray._raylet.execute_task.function_executor
File "D:\Anaconda\envs\time_series_py38\lib\site-packages\ray_private\function_manager.py", line 674, in actor_method_executor
return method(__ray_actor, *args, **kwargs)
File "D:\Anaconda\envs\time_series_py38\lib\site-packages\ray\util\tracing\tracing_helper.py", line 466, in _resume_span
return method(self, *_args, **_kwargs)
File "D:\Anaconda\envs\time_series_py38\lib\site-packages\ray\tune\trainable\trainable.py", line 355, in train
raise skipped from exception_cause(skipped)
File "D:\Anaconda\envs\time_series_py38\lib\site-packages\ray\tune\trainable\function_trainable.py", line 325, in entrypoint
return self._trainable_func(
File "D:\Anaconda\envs\time_series_py38\lib\site-packages\ray\util\tracing\tracing_helper.py", line 466, in _resume_span
return method(self, *_args, **_kwargs)
File "D:\Anaconda\envs\time_series_py38\lib\site-packages\ray\tune\trainable\function_trainable.py", line 651, in _trainable_func
output = fn()
File "C:\Users\ouyeel\AppData\Roaming\Python\Python38\site-packages\paddlets\automl\optimize_runner.py", line 174, in run_trial
score = fit_and_score(train_data=train_tsdataset,
File "C:\Users\ouyeel\AppData\Roaming\Python\Python38\site-packages\paddlets\utils\validation.py", line 135, in fit_and_score
and check_train_valid_continuity(train_data, valid_data):
File "C:\Users\ouyeel\AppData\Roaming\Python\Python38\site-packages\paddlets\utils\utils.py", line 119, in check_train_valid_continuity
continuious = (valid_index[0] - train_index[-1] == pd.to_timedelta(train_index.freq))
File "C:\Users\ouyeel\AppData\Roaming\Python\Python38\site-packages\pandas\core\tools\timedeltas.py", line 142, in to_timedelta
return _coerce_scalar_to_timedelta_type(arg, unit=unit, errors=errors)
File "C:\Users\ouyeel\AppData\Roaming\Python\Python38\site-packages\pandas\core\tools\timedeltas.py", line 150, in _coerce_scalar_to_timedelta_type
result = Timedelta(r, unit)
File "pandas_libs\tslibs\timedeltas.pyx", line 1315, in pandas._libs.tslibs.timedeltas.Timedelta.new
ValueError: Value must be Timedelta, string, integer, float, timedelta or convertible, not MonthEnd

代码如下：
from paddlets.utils import backtest

q_loss, quantiles = backtest(data=ts_test_scaled,
model=deepar,
start="2021/9/24 19:00",
metric=QuantileLoss([0.1, 0.5, 0.9]),
predict_window=306,
stride=306,
return_predicts=True
)

quantiles.plot(
add_data=ts_test_scaled,
low_quantile=0.05,
high_quantile=0.95
)
plt.show()
报错如下：
[2022-11-07 16:10:23,106] [paddlets.models.common.callbacks.callbacks] [INFO] Best weights from best epoch are automatically used!
[2022-11-07 16:10:23,119] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:23,120] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:23,120] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:23,120] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:23,120] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:23,120] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:23,120] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:23,120] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:23,155] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:23,207] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:23,217] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:23,227] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:23,241] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:23,262] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:23,280] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:23,300] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,822] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,822] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,822] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,823] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,823] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,823] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,823] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,823] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,885] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,905] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,908] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,934] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,948] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,954] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,956] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,956] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,963] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,963] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,964] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,964] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,968] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:31,968] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:32,000] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:32,004] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:32,034] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:32,080] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:32,084] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:32,120] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:32,123] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:32,142] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:32,159] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:32,180] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:34,758] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:34,758] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:34,758] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:34,758] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:34,758] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:34,759] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:34,759] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:34,759] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:34,817] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:34,821] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:34,844] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:34,861] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:34,911] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:34,913] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:34,917] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:34,935] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:35,653] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:35,653] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:35,653] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:35,653] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:35,654] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:35,654] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:35,654] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:35,654] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:35,736] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:35,748] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:35,754] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:35,756] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:35,795] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:35,797] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:35,814] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:35,831] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:36,401] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:36,401] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:36,401] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:36,401] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:36,401] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:36,402] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:36,402] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:36,402] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:36,437] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:36,464] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:36,501] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:36,504] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:36,538] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:36,541] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:36,560] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:36,578] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,085] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,085] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,085] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,085] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,085] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,086] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,086] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,086] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,133] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,167] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,184] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,190] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,241] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,245] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,248] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,264] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,746] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,746] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,746] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,746] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,747] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,747] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,747] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,747] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,788] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,828] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,846] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,863] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,871] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,887] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,907] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:37,923] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:38,344] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:38,344] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:38,344] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:38,344] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:38,344] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:38,345] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:38,345] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:38,345] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:38,387] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:38,423] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:38,430] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:38,447] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:38,467] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:38,499] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:38,503] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:38,521] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,035] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,035] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,035] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,035] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,035] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,036] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,036] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,036] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,086] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,123] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,137] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,139] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,139] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,142] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,144] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,145] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,146] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,146] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,147] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,148] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,168] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,199] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,217] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,221] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,244] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,286] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,306] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,311] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,327] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,347] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,367] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:39,383] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
C:\ProgramData\Anaconda3\envs\paddlets\lib\site-packages\paddlets\automl\searcher.py:4: DeprecationWarning: The module ray.tune.suggest has been moved to ray.tune.search and the old location will be deprecated soon. Please adjust your imports to point to the new location. Example: Do a global search and replace ray.tune.suggest with ray.tune.search.
from ray.tune.suggest import BasicVariantGenerator
C:\ProgramData\Anaconda3\envs\paddlets\lib\site-packages\paddlets\automl\searcher.py:5: DeprecationWarning: The module ray.tune.suggest.optuna has been moved to ray.tune.search.optuna and the old location will be deprecated soon. Please adjust your imports to point to the new location. Example: Do a global search and replace ray.tune.suggest.optuna with ray.tune.search.optuna.
from ray.tune.suggest.optuna import OptunaSearch
C:\ProgramData\Anaconda3\envs\paddlets\lib\site-packages\flaml\tune_init_.py:5: DeprecationWarning: The module ray.tune.sample has been moved to ray.tune.search.sample and the old location will be deprecated soon. Please adjust your imports to point to the new location. Example: Do a global search and replace ray.tune.sample with ray.tune.search.sample.
from ray.tune import (
C:\ProgramData\Anaconda3\envs\paddlets\lib\site-packages\flaml\tune\space.py:6: DeprecationWarning: The module ray.tune.suggest.variant_generator has been moved to ray.tune.search.variant_generator and the old location will be deprecated soon. Please adjust your imports to point to the new location. Example: Do a global search and replace ray.tune.suggest.variant_generator with ray.tune.search.variant_generator.
from ray.tune.suggest.variant_generator import generate_variants
C:\ProgramData\Anaconda3\envs\paddlets\lib\site-packages\paddlets\automl\searcher.py:6: DeprecationWarning: The module ray.tune.suggest.flaml has been moved to ray.tune.search.flaml and the old location will be deprecated soon. Please adjust your imports to point to the new location. Example: Do a global search and replace ray.tune.suggest.flaml with ray.tune.search.flaml.
from ray.tune.suggest.flaml import CFO
C:\ProgramData\Anaconda3\envs\paddlets\lib\site-packages\paddlets\automl\searcher.py:8: DeprecationWarning: The module ray.tune.suggest.bohb has been moved to ray.tune.search.bohb and the old location will be deprecated soon. Please adjust your imports to point to the new location. Example: Do a global search and replace ray.tune.suggest.bohb with ray.tune.search.bohb.
from ray.tune.suggest.bohb import TuneBOHB
Backtest Progress: 0%| | 0/9 [00:00<?, ?it/s]C:\ProgramData\Anaconda3\envs\paddlets\lib\site-packages\paddle\fluid\layers\tensor.py:657: UserWarning: paddle.assign doesn't support float64 input now due to current platform protobuf data limitation, we convert it to float32
warnings.warn(
Backtest Progress: 0%| | 0/9 [00:01<?, ?it/s]
[2022-11-07 16:10:57,802] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:57,802] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:57,803] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:57,803] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:57,803] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:57,803] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:57,803] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:57,803] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:57,843] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:57,867] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:57,902] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:57,904] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:57,953] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:57,959] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:57,962] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-11-07 16:10:57,983] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
Traceback (most recent call last):
File "C:\Program Files\JetBrains\PyCharm 2021.3.2\plugins\python\helpers\pydev\pydevd.py", line 1483, in _exec
pydev_imports.execfile(file, globals, locals) # execute the script
File "C:\Program Files\JetBrains\PyCharm 2021.3.2\plugins\python\helpers\pydev_pydev_imps_pydev_execfile.py", line 18, in execfile
exec(compile(contents+"\n", file, 'exec'), glob, loc)
File "E:/Paddle-release-2.2/PaddleTS/demos/fit_deepar_model.py", line 46, in
q_loss, quantiles = backtest(data=ts_test_scaled,
File "C:\ProgramData\Anaconda3\envs\paddlets\lib\site-packages\paddlets\utils\backtest.py", line 150, in backtest
score_dict = metric(real, predict)
File "C:\ProgramData\Anaconda3\envs\paddlets\lib\site-packages\paddlets\metrics\base.py", line 262, in call
res_array = self._build_prob_metrics_data(tsdataset_true, tsdataset_pred, self._TYPE)
File "C:\ProgramData\Anaconda3\envs\paddlets\lib\site-packages\paddlets\metrics\base.py", line 104, in _build_prob_metrics_data
target_true, target_pred = self._reindex_data(target_true, target_pred)
File "C:\ProgramData\Anaconda3\envs\paddlets\lib\site-packages\paddlets\metrics\base.py", line 197, in _reindex_data
merge_index = pd.merge(target_true.time_index.to_frame(index=False), target_pred.time_index.to_frame(index=False))
File "C:\ProgramData\Anaconda3\envs\paddlets\lib\site-packages\pandas\core\reshape\merge.py", line 106, in merge
op = _MergeOperation(
File "C:\ProgramData\Anaconda3\envs\paddlets\lib\site-packages\pandas\core\reshape\merge.py", line 681, in init
self._validate_specification()
File "C:\ProgramData\Anaconda3\envs\paddlets\lib\site-packages\pandas\core\reshape\merge.py", line 1346, in _validate_specification
raise MergeError(
pandas.errors.MergeError: No common columns to perform merge on. Merge options: left_on=None, right_on=None, left_index=False, right_index=False

运行文档https://github.com/PaddlePaddle/PaddleTS/blob/main/examples/autots_example.ipynb到autots_model.fit(tsdataset)这一步的时候出现以下报错：
(pid=30292) C:\Users\Administrator\AppData\Roaming\Python\Python38\site-packages\paddlets\automl\searcher.py:4: DeprecationWarning: The module ray.tune.suggest has been moved to ray.tune.search and the old location will be deprecated soon. Please adjust your imports to point to the new location. Example: Do a global search and replace ray.tune.suggest with ray.tune.search.
(pid=30292) from ray.tune.suggest import BasicVariantGenerator
(pid=30292) C:\Users\Administrator\AppData\Roaming\Python\Python38\site-packages\paddlets\automl\searcher.py:5: DeprecationWarning: The module ray.tune.suggest.optuna has been moved to ray.tune.search.optuna and the old location will be deprecated soon. Please adjust your imports to point to the new location. Example: Do a global search and replace ray.tune.suggest.optuna with ray.tune.search.optuna.
(pid=30292) from ray.tune.suggest.optuna import OptunaSearch
(pid=30292) C:\Users\Administrator\AppData\Roaming\Python\Python38\site-packages\paddlets\automl\searcher.py:6: DeprecationWarning: The module ray.tune.suggest.flaml has been moved to ray.tune.search.flaml and the old location will be deprecated soon. Please adjust your imports to point to the new location. Example: Do a global search and replace ray.tune.suggest.flaml with ray.tune.search.flaml.
(pid=30292) from ray.tune.suggest.flaml import CFO
(pid=30292) C:\Users\Administrator\AppData\Roaming\Python\Python38\site-packages\paddlets\automl\searcher.py:8: DeprecationWarning: The module ray.tune.suggest.bohb has been moved to ray.tune.search.bohb and the old location will be deprecated soon. Please adjust your imports to point to the new location. Example: Do a global search and replace ray.tune.suggest.bohb with ray.tune.search.bohb.
(pid=30292) from ray.tune.suggest.bohb import TuneBOHB
(pid=30292) C:\Users\Administrator\AppData\Roaming\Python\Python38\site-packages\paddlets\automl\search_space_configer.py:8: DeprecationWarning: The module ray.tune.sample has been moved to ray.tune.search.sample and the old location will be deprecated soon. Please adjust your imports to point to the new location. Example: Do a global search and replace ray.tune.sample with ray.tune.search.sample.
(pid=30292) from ray.tune.sample import Float, Integer, Categorical
(run_trial pid=30292) [2022-11-14 01:01:15,022] [paddlets.automl.optimize_runner] [INFO] trial config: {'hidden_config': 'Choice_0: [64]', 'use_bn': True, 'batch_size': 112, 'max_epochs': 330, 'optimizer_params': {'learning_rate': 0.009473699093873009}, 'patience': 15}
(run_trial pid=30292) [2022-11-14 01:01:15,022] [paddlets.automl.optimize_runner] [INFO] setup_estimator: init model. Params: {'hidden_config': [64], 'use_bn': True, 'batch_size': 112, 'max_epochs': 330, 'optimizer_params': {'learning_rate': 0.009473699093873009}, 'patience': 15, 'in_chunk_len': 96, 'out_chunk_len': 2, 'skip_chunk_len': 0, 'sampling_stride': 1, 'seed': 2022}
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(

KeyError Traceback (most recent call last)
~\anaconda3\lib\site-packages\tensorboardX\record_writer.py in open_file(path)
57 prefix = path.split(':')[0]
---> 58 factory = REGISTERED_FACTORIES[prefix]
59 return factory.open(path)

KeyError: 'C'

During handling of the above exception, another exception occurred:

FileNotFoundError Traceback (most recent call last)
~\anaconda3\lib\site-packages\ray\tune\execution\trial_runner.py in _wait_and_handle_event(self, next_trial)
832 if event.type == _ExecutorEventType.PG_READY:
--> 833 self._on_pg_ready(next_trial)
834 elif event.type == _ExecutorEventType.NO_RUNNING_TRIAL_TIMEOUT:

~\anaconda3\lib\site-packages\ray\tune\execution\trial_runner.py in _on_pg_ready(self, next_trial)
922 logger.debug(f"Trying to start trial: {next_trial}")
--> 923 if not _start_trial(next_trial) and next_trial.status != Trial.ERROR:
924 # Only try to start another trial if previous trial startup

~\anaconda3\lib\site-packages\ray\tune\execution\trial_runner.py in _start_trial(trial)
914 if self.trial_executor.start_trial(trial):
--> 915 self._callbacks.on_trial_start(
916 iteration=self._iteration, trials=self._trials, trial=trial

~\anaconda3\lib\site-packages\ray\tune\callback.py in on_trial_start(self, **info)
316 for callback in self._callbacks:
--> 317 callback.on_trial_start(**info)
318

~\anaconda3\lib\site-packages\ray\tune\logger\logger.py in on_trial_start(self, iteration, trials, trial, **info)
134 ):
--> 135 self.log_trial_start(trial)
136

~\anaconda3\lib\site-packages\ray\tune\logger\tensorboardx.py in log_trial_start(self, trial)
178 trial.init_logdir()
--> 179 self._trial_writer[trial] = self._summary_writer_cls(
180 trial.logdir, flush_secs=30

~\anaconda3\lib\site-packages\tensorboardX\writer.py in init(self, logdir, comment, purge_step, max_queue, flush_secs, filename_suffix, write_to_disk, log_dir, comet_config, **kwargs)
300 self.file_writer = self.all_writers = None
--> 301 self._get_file_writer()
302

~\anaconda3\lib\site-packages\tensorboardX\writer.py in _get_file_writer(self)
348 if self.all_writers is None or self.file_writer is None:
--> 349 self.file_writer = FileWriter(logdir=self.logdir,
350 max_queue=self._max_queue,

~\anaconda3\lib\site-packages\tensorboardX\writer.py in init(self, logdir, max_queue, flush_secs, filename_suffix)
104 logdir = str(logdir)
--> 105 self.event_writer = EventFileWriter(
106 logdir, max_queue, flush_secs, filename_suffix)

~\anaconda3\lib\site-packages\tensorboardX\event_file_writer.py in init(self, logdir, max_queue_size, flush_secs, filename_suffix)
105 self._event_queue = multiprocessing.Queue(max_queue_size)
--> 106 self._ev_writer = EventsWriter(os.path.join(
107 self._logdir, "events"), filename_suffix)

~\anaconda3\lib\site-packages\tensorboardX\event_file_writer.py in init(self, file_prefix, filename_suffix)
42 self._num_outstanding_events = 0
---> 43 self._py_recordio_writer = RecordWriter(self._file_name)
44 # Initialize an event instance.

~\anaconda3\lib\site-packages\tensorboardX\record_writer.py in init(self, path)
178 self._writer = None
--> 179 self._writer = open_file(path)
180

~\anaconda3\lib\site-packages\tensorboardX\record_writer.py in open_file(path)
60 except KeyError:
---> 61 return open(path, 'wb')
62

FileNotFoundError: [Errno 2] No such file or directory: 'C:\Users\Administrator\ray_results\run_trial_2022-11-14_01-01-08\run_trial_c88b9bca_2_batch_size=120,hidden_config=Choice_5_128_128_128,max_epochs=180,learning_rate=0.0042,patience=45,use_bn=Fals_2022-11-14_01-01-15\events.out.tfevents.1668358875.SK-20200101ELSU'

During handling of the above exception, another exception occurred:

TuneError Traceback (most recent call last)
in
----> 1 autots_model.fit(tsdataset)

~\AppData\Roaming\Python\Python38\site-packages\paddlets\logger\logger.py in wrapper(*args, **kwargs)
25 logger = Logger(module_name)
26 logger.debug("function:%s" % func_name)
---> 27 result = f(*args, **kwargs)
28 return result
29

~\AppData\Roaming\Python\Python38\site-packages\paddlets\automl\autots.py in fit(self, train_tsdataset, valid_tsdataset, n_trails, cpu_resource, gpu_resource)
205 if valid_tsdataset is None:
206 raise NotImplementedError("When the train_tsdataset is a list, valid_tsdataset is required!")
--> 207 analysis = self._optimize_runner.optimize(self._estimator,
208 self._in_chunk_len,
209 self._out_chunk_len,

~\AppData\Roaming\Python\Python38\site-packages\paddlets\automl\optimize_runner.py in optimize(self, paddlets_estimator, in_chunk_len, out_chunk_len, train_tsdataset, valid_tsdataset, sampling_stride, skip_chunk_len, search_space, metric, mode, resampling_strategy, split_ratio, k_fold, n_trials, cpu_resource, gpu_resource, local_dir)
202 self._track_choice_mapping = self._preprocess_search_space(running_search_space)
203
--> 204 return tune.run(run_trial, num_samples=n_trials, config=running_search_space,
205 metric=self.report_metric,
206 mode=mode,

~\anaconda3\lib\site-packages\ray\tune\tune.py in run(run_or_experiment, name, metric, mode, stop, time_budget_s, config, resources_per_trial, num_samples, local_dir, search_alg, scheduler, keep_checkpoints_num, checkpoint_score_attr, checkpoint_freq, checkpoint_at_end, verbose, progress_reporter, log_to_file, trial_name_creator, trial_dirname_creator, sync_config, export_formats, max_failures, fail_fast, restore, server_port, resume, reuse_actors, trial_executor, raise_on_failed_trial, callbacks, max_concurrent_trials, _experiment_checkpoint_dir, _remote)
739 )
740 while not runner.is_finished() and not state["signal"]:
--> 741 runner.step()
742 if has_verbosity(Verbosity.V1_EXPERIMENT):
743 _report_progress(runner, progress_reporter)

~\anaconda3\lib\site-packages\ray\tune\execution\trial_runner.py in step(self)
884 logger.debug(f"Got new trial to run: {next_trial}")
885
--> 886 self._wait_and_handle_event(next_trial)
887
888 self._stop_experiment_if_needed()

~\anaconda3\lib\site-packages\ray\tune\execution\trial_runner.py in _wait_and_handle_event(self, next_trial)
863 raise e
864 else:
--> 865 raise TuneError(traceback.format_exc())
866
867 def step(self):

TuneError: Traceback (most recent call last):
File "C:\Users\Administrator\anaconda3\lib\site-packages\tensorboardX\record_writer.py", line 58, in open_file
factory = REGISTERED_FACTORIES[prefix]
KeyError: 'C'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "C:\Users\Administrator\anaconda3\lib\site-packages\ray\tune\execution\trial_runner.py", line 833, in _wait_and_handle_event
self._on_pg_ready(next_trial)
File "C:\Users\Administrator\anaconda3\lib\site-packages\ray\tune\execution\trial_runner.py", line 923, in _on_pg_ready
if not _start_trial(next_trial) and next_trial.status != Trial.ERROR:
File "C:\Users\Administrator\anaconda3\lib\site-packages\ray\tune\execution\trial_runner.py", line 915, in _start_trial
self._callbacks.on_trial_start(
File "C:\Users\Administrator\anaconda3\lib\site-packages\ray\tune\callback.py", line 317, in on_trial_start
callback.on_trial_start(**info)
File "C:\Users\Administrator\anaconda3\lib\site-packages\ray\tune\logger\logger.py", line 135, in on_trial_start
self.log_trial_start(trial)
File "C:\Users\Administrator\anaconda3\lib\site-packages\ray\tune\logger\tensorboardx.py", line 179, in log_trial_start
self._trial_writer[trial] = self._summary_writer_cls(
File "C:\Users\Administrator\anaconda3\lib\site-packages\tensorboardX\writer.py", line 301, in init
self._get_file_writer()
File "C:\Users\Administrator\anaconda3\lib\site-packages\tensorboardX\writer.py", line 349, in _get_file_writer
self.file_writer = FileWriter(logdir=self.logdir,
File "C:\Users\Administrator\anaconda3\lib\site-packages\tensorboardX\writer.py", line 105, in init
self.event_writer = EventFileWriter(
File "C:\Users\Administrator\anaconda3\lib\site-packages\tensorboardX\event_file_writer.py", line 106, in init
self._ev_writer = EventsWriter(os.path.join(
File "C:\Users\Administrator\anaconda3\lib\site-packages\tensorboardX\event_file_writer.py", line 43, in init
self._py_recordio_writer = RecordWriter(self._file_name)
File "C:\Users\Administrator\anaconda3\lib\site-packages\tensorboardX\record_writer.py", line 179, in init
self._writer = open_file(path)
File "C:\Users\Administrator\anaconda3\lib\site-packages\tensorboardX\record_writer.py", line 61, in open_file
return open(path, 'wb')
FileNotFoundError: [Errno 2] No such file or directory: 'C:\Users\Administrator\ray_results\run_trial_2022-11-14_01-01-08\run_trial_c88b9bca_2_batch_size=120,hidden_config=Choice_5_128_128_128,max_epochs=180,learning_rate=0.0042,patience=45,use_bn=Fals_2022-11-14_01-01-15\events.out.tfevents.1668358875.SK-20200101ELSU'

(run_trial pid=30292) [2022-11-14 01:01:15,572] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 000| loss: 3.339083| val_0_mae: 0.883445| 0:00:00s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=30292) [2022-11-14 01:01:15,993] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 001| loss: 1.339651| val_0_mae: 0.780203| 0:00:00s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=30292) [2022-11-14 01:01:16,418] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 002| loss: 1.073304| val_0_mae: 0.742995| 0:00:01s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=30292) [2022-11-14 01:01:16,830] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 003| loss: 1.000667| val_0_mae: 0.761179| 0:00:01s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=30292) [2022-11-14 01:01:17,249] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 004| loss: 0.964694| val_0_mae: 0.707594| 0:00:02s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=30292) [2022-11-14 01:01:17,678] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 005| loss: 0.972320| val_0_mae: 0.850871| 0:00:02s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=30292) [2022-11-14 01:01:18,108] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 006| loss: 0.966438| val_0_mae: 0.885582| 0:00:02s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=30292) [2022-11-14 01:01:18,520] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 007| loss: 0.896922| val_0_mae: 0.909875| 0:00:03s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=30292) [2022-11-14 01:01:18,943] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 008| loss: 0.932748| val_0_mae: 0.702189| 0:00:03s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=30292) [2022-11-14 01:01:19,355] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 009| loss: 0.883199| val_0_mae: 0.681072| 0:00:04s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(pid=42476) C:\Users\Administrator\AppData\Roaming\Python\Python38\site-packages\paddlets\automl\searcher.py:4: DeprecationWarning: The module ray.tune.suggest has been moved to ray.tune.search and the old location will be deprecated soon. Please adjust your imports to point to the new location. Example: Do a global search and replace ray.tune.suggest with ray.tune.search.
(pid=42476) from ray.tune.suggest import BasicVariantGenerator
(run_trial pid=30292) [2022-11-14 01:01:19,774] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 010| loss: 0.942490| val_0_mae: 0.726340| 0:00:04s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(pid=42476) C:\Users\Administrator\AppData\Roaming\Python\Python38\site-packages\paddlets\automl\searcher.py:5: DeprecationWarning: The module ray.tune.suggest.optuna has been moved to ray.tune.search.optuna and the old location will be deprecated soon. Please adjust your imports to point to the new location. Example: Do a global search and replace ray.tune.suggest.optuna with ray.tune.search.optuna.
(pid=42476) from ray.tune.suggest.optuna import OptunaSearch
(run_trial pid=30292) [2022-11-14 01:01:20,197] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 011| loss: 0.863567| val_0_mae: 0.679934| 0:00:05s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(pid=42476) C:\Users\Administrator\AppData\Roaming\Python\Python38\site-packages\paddlets\automl\searcher.py:6: DeprecationWarning: The module ray.tune.suggest.flaml has been moved to ray.tune.search.flaml and the old location will be deprecated soon. Please adjust your imports to point to the new location. Example: Do a global search and replace ray.tune.suggest.flaml with ray.tune.search.flaml.
(pid=42476) from ray.tune.suggest.flaml import CFO
(run_trial pid=30292) [2022-11-14 01:01:20,627] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 012| loss: 0.920713| val_0_mae: 0.684235| 0:00:05s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(pid=42476) C:\Users\Administrator\AppData\Roaming\Python\Python38\site-packages\paddlets\automl\searcher.py:8: DeprecationWarning: The module ray.tune.suggest.bohb has been moved to ray.tune.search.bohb and the old location will be deprecated soon. Please adjust your imports to point to the new location. Example: Do a global search and replace ray.tune.suggest.bohb with ray.tune.search.bohb.
(pid=42476) from ray.tune.suggest.bohb import TuneBOHB
(pid=42476) C:\Users\Administrator\AppData\Roaming\Python\Python38\site-packages\paddlets\automl\search_space_configer.py:8: DeprecationWarning: The module ray.tune.sample has been moved to ray.tune.search.sample and the old location will be deprecated soon. Please adjust your imports to point to the new location. Example: Do a global search and replace ray.tune.sample with ray.tune.search.sample.
(pid=42476) from ray.tune.sample import Float, Integer, Categorical
(run_trial pid=30292) [2022-11-14 01:01:21,056] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 013| loss: 0.847748| val_0_mae: 0.739015| 0:00:05s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=42476) [2022-11-14 01:01:21,147] [paddlets.automl.optimize_runner] [INFO] trial config: {'hidden_config': 'Choice_5: [128, 128, 128]', 'use_bn': False, 'batch_size': 120, 'max_epochs': 180, 'optimizer_params': {'learning_rate': 0.004155371012314867}, 'patience': 45}
(run_trial pid=42476) [2022-11-14 01:01:21,147] [paddlets.automl.optimize_runner] [INFO] setup_estimator: init model. Params: {'hidden_config': [128, 128, 128], 'use_bn': False, 'batch_size': 120, 'max_epochs': 180, 'optimizer_params': {'learning_rate': 0.004155371012314867}, 'patience': 45, 'in_chunk_len': 96, 'out_chunk_len': 2, 'skip_chunk_len': 0, 'sampling_stride': 1, 'seed': 2022}
(run_trial pid=30292) [2022-11-14 01:01:21,477] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 014| loss: 0.877982| val_0_mae: 0.915531| 0:00:06s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=42476) [2022-11-14 01:01:21,839] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 000| loss: 2.250421| val_0_mae: 0.866377| 0:00:00s
(run_trial pid=30292) [2022-11-14 01:01:21,898] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 015| loss: 0.865720| val_0_mae: 1.103737| 0:00:06s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=30292) [2022-11-14 01:01:22,317] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 016| loss: 0.882259| val_0_mae: 0.690138| 0:00:07s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=42476) [2022-11-14 01:01:22,436] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 001| loss: 1.000779| val_0_mae: 0.903012| 0:00:01s
(run_trial pid=30292) [2022-11-14 01:01:22,734] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 017| loss: 0.853780| val_0_mae: 0.726888| 0:00:07s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=42476) [2022-11-14 01:01:23,041] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 002| loss: 0.970277| val_0_mae: 0.745609| 0:00:01s
(run_trial pid=30292) [2022-11-14 01:01:23,157] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 018| loss: 0.864675| val_0_mae: 0.746551| 0:00:08s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=30292) [2022-11-14 01:01:23,573] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 019| loss: 0.852602| val_0_mae: 0.806146| 0:00:08s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=42476) [2022-11-14 01:01:23,761] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 003| loss: 0.924974| val_0_mae: 0.824449| 0:00:02s
(run_trial pid=30292) [2022-11-14 01:01:23,990] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 020| loss: 0.865885| val_0_mae: 0.668478| 0:00:08s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=30292) [2022-11-14 01:01:24,421] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 021| loss: 0.822864| val_0_mae: 0.782634| 0:00:09s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=42476) [2022-11-14 01:01:24,499] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 004| loss: 0.923414| val_0_mae: 0.789935| 0:00:03s
(run_trial pid=30292) [2022-11-14 01:01:24,829] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 022| loss: 0.854838| val_0_mae: 0.668849| 0:00:09s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=30292) [2022-11-14 01:01:25,238] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 023| loss: 0.832975| val_0_mae: 0.848204| 0:00:10s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=42476) [2022-11-14 01:01:25,267] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 005| loss: 0.884985| val_0_mae: 0.797978| 0:00:04s
(run_trial pid=30292) [2022-11-14 01:01:25,646] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 024| loss: 0.803786| val_0_mae: 0.771176| 0:00:10s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=30292) [2022-11-14 01:01:26,063] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 025| loss: 0.831367| val_0_mae: 0.755703| 0:00:10s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=42476) [2022-11-14 01:01:26,054] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 006| loss: 0.846988| val_0_mae: 0.720191| 0:00:04s
(run_trial pid=30292) [2022-11-14 01:01:26,473] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 026| loss: 0.809912| val_0_mae: 0.741790| 0:00:11s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=30292) [2022-11-14 01:01:26,881] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 027| loss: 0.813126| val_0_mae: 0.743538| 0:00:11s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=42476) [2022-11-14 01:01:26,853] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 007| loss: 0.790365| val_0_mae: 0.783437| 0:00:05s
(run_trial pid=30292) [2022-11-14 01:01:27,289] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 028| loss: 0.798136| val_0_mae: 0.682494| 0:00:12s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=42476) [2022-11-14 01:01:27,680] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 008| loss: 0.816146| val_0_mae: 0.837121| 0:00:06s
(run_trial pid=30292) [2022-11-14 01:01:27,698] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 029| loss: 0.816031| val_0_mae: 0.689467| 0:00:12s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=30292) [2022-11-14 01:01:28,115] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 030| loss: 0.798248| val_0_mae: 0.810028| 0:00:13s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=42476) [2022-11-14 01:01:28,521] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 009| loss: 0.844155| val_0_mae: 0.773559| 0:00:07s
(run_trial pid=30292) [2022-11-14 01:01:28,536] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 031| loss: 0.782819| val_0_mae: 0.675081| 0:00:13s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=30292) [2022-11-14 01:01:28,941] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 032| loss: 0.830348| val_0_mae: 0.703546| 0:00:13s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=30292) [2022-11-14 01:01:29,344] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 033| loss: 0.787820| val_0_mae: 0.671389| 0:00:14s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=42476) [2022-11-14 01:01:29,361] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 010| loss: 0.800365| val_0_mae: 0.769280| 0:00:08s
(run_trial pid=30292) [2022-11-14 01:01:29,764] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 034| loss: 0.812075| val_0_mae: 0.731617| 0:00:14s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\paddle\nn\layer\norm.py:653: UserWarning: When training, we now always track global mean and variance.
(run_trial pid=30292) warnings.warn(
(run_trial pid=30292) [2022-11-14 01:01:30,184] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 035| loss: 0.808069| val_0_mae: 0.674505| 0:00:15s
(run_trial pid=30292) [2022-11-14 01:01:30,185] [paddlets.models.common.callbacks.callbacks] [INFO]
(run_trial pid=30292) Early stopping occurred at epoch 35 with best_epoch = 20 and best_val_0_mae = 0.668478
(run_trial pid=30292) [2022-11-14 01:01:30,185] [paddlets.models.common.callbacks.callbacks] [INFO] Best weights from best epoch are automatically used!
(run_trial pid=42476) [2022-11-14 01:01:30,183] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 011| loss: 0.814131| val_0_mae: 0.712737| 0:00:08s
(run_trial pid=42476) [2022-11-14 01:01:31,032] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 012| loss: 0.808135| val_0_mae: 0.739639| 0:00:09s
(run_trial pid=42476) [2022-11-14 01:01:31,897] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 013| loss: 0.814179| val_0_mae: 0.755509| 0:00:10s
(run_trial pid=42476) [2022-11-14 01:01:32,795] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 014| loss: 0.815168| val_0_mae: 0.766864| 0:00:11s
(run_trial pid=42476) [2022-11-14 01:01:33,732] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 015| loss: 0.753962| val_0_mae: 0.690585| 0:00:12s
(run_trial pid=42476) [2022-11-14 01:01:34,675] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 016| loss: 0.801936| val_0_mae: 0.709537| 0:00:13s
(run_trial pid=42476) [2022-11-14 01:01:35,645] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 017| loss: 0.794926| val_0_mae: 0.749197| 0:00:14s
(run_trial pid=42476) [2022-11-14 01:01:36,600] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 018| loss: 0.762111| val_0_mae: 0.678700| 0:00:15s
(run_trial pid=42476) [2022-11-14 01:01:37,540] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 019| loss: 0.745253| val_0_mae: 0.922469| 0:00:16s
(run_trial pid=42476) [2022-11-14 01:01:38,506] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 020| loss: 0.787604| val_0_mae: 0.700718| 0:00:17s
(run_trial pid=42476) [2022-11-14 01:01:39,488] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 021| loss: 0.732736| val_0_mae: 0.741522| 0:00:18s
(run_trial pid=42476) [2022-11-14 01:01:40,478] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 022| loss: 0.756187| val_0_mae: 0.687385| 0:00:19s
(run_trial pid=42476) [2022-11-14 01:01:41,501] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 023| loss: 0.751891| val_0_mae: 0.690699| 0:00:20s
(run_trial pid=42476) [2022-11-14 01:01:42,519] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 024| loss: 0.727448| val_0_mae: 0.682869| 0:00:21s
(run_trial pid=42476) [2022-11-14 01:01:43,555] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 025| loss: 0.755604| val_0_mae: 0.895392| 0:00:22s
(run_trial pid=42476) [2022-11-14 01:01:44,597] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 026| loss: 0.792194| val_0_mae: 0.706429| 0:00:23s
(run_trial pid=42476) [2022-11-14 01:01:45,655] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 027| loss: 0.744930| val_0_mae: 0.791593| 0:00:24s
(run_trial pid=30292) C:\Users\Administrator\anaconda3\lib\site-packages\ray\tune\trainable\session.py:229: DeprecationWarning: tune.report and tune.checkpoint_dir APIs are deprecated in Ray 2.0, and is replaced by ray.air.session. This will provide an easy-to-use API across Tune session and Data parallel worker sessions.The old APIs will be removed in the future.
(run_trial pid=30292) warnings.warn(
(run_trial pid=42476) [2022-11-14 01:01:46,718] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 028| loss: 0.748070| val_0_mae: 0.791396| 0:00:25s
(run_trial pid=42476) [2022-11-14 01:01:47,765] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 029| loss: 0.728586| val_0_mae: 0.787615| 0:00:26s
(run_trial pid=42476) [2022-11-14 01:01:48,822] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 030| loss: 0.729428| val_0_mae: 0.750572| 0:00:27s
(run_trial pid=42476) [2022-11-14 01:01:49,869] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 031| loss: 0.711299| val_0_mae: 0.708207| 0:00:28s
(run_trial pid=42476) [2022-11-14 01:01:50,932] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 032| loss: 0.735087| val_0_mae: 0.711936| 0:00:29s
(run_trial pid=42476) [2022-11-14 01:01:52,010] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 033| loss: 0.730024| val_0_mae: 0.741676| 0:00:30s
(run_trial pid=42476) [2022-11-14 01:01:53,085] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 034| loss: 0.703183| val_0_mae: 0.755382| 0:00:31s
(run_trial pid=42476) [2022-11-14 01:01:54,170] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 035| loss: 0.704181| val_0_mae: 0.686618| 0:00:32s
(run_trial pid=42476) [2022-11-14 01:01:55,267] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 036| loss: 0.704718| val_0_mae: 0.724146| 0:00:34s
(run_trial pid=42476) [2022-11-14 01:01:56,361] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 037| loss: 0.693106| val_0_mae: 0.697631| 0:00:35s
(run_trial pid=42476) [2022-11-14 01:01:57,466] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 038| loss: 0.727850| val_0_mae: 0.662641| 0:00:36s
(run_trial pid=42476) [2022-11-14 01:01:58,583] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 039| loss: 0.688710| val_0_mae: 0.698511| 0:00:37s
(run_trial pid=42476) [2022-11-14 01:01:59,692] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 040| loss: 0.685247| val_0_mae: 0.702356| 0:00:38s
(run_trial pid=42476) [2022-11-14 01:02:00,810] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 041| loss: 0.702037| val_0_mae: 0.677072| 0:00:39s
(run_trial pid=42476) [2022-11-14 01:02:01,935] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 042| loss: 0.675219| val_0_mae: 0.693985| 0:00:40s
(run_trial pid=42476) [2022-11-14 01:02:03,079] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 043| loss: 0.665016| val_0_mae: 0.688647| 0:00:41s
(run_trial pid=42476) [2022-11-14 01:02:04,227] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 044| loss: 0.714889| val_0_mae: 0.741173| 0:00:42s
(run_trial pid=42476) [2022-11-14 01:02:05,384] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 045| loss: 0.692830| val_0_mae: 0.690192| 0:00:44s
(run_trial pid=42476) [2022-11-14 01:02:06,557] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 046| loss: 0.674419| val_0_mae: 0.760584| 0:00:45s
(run_trial pid=42476) [2022-11-14 01:02:07,734] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 047| loss: 0.694529| val_0_mae: 0.695555| 0:00:46s
(run_trial pid=42476) [2022-11-14 01:02:08,914] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 048| loss: 0.670605| val_0_mae: 0.821415| 0:00:47s
(run_trial pid=42476) [2022-11-14 01:02:10,092] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 049| loss: 0.666438| val_0_mae: 0.755908| 0:00:48s
(run_trial pid=42476) [2022-11-14 01:02:11,287] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 050| loss: 0.699248| val_0_mae: 0.679514| 0:00:50s
(run_trial pid=42476) [2022-11-14 01:02:12,483] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 051| loss: 0.636179| val_0_mae: 0.672630| 0:00:51s
(run_trial pid=42476) [2022-11-14 01:02:13,704] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 052| loss: 0.654464| val_0_mae: 0.702088| 0:00:52s
(run_trial pid=42476) [2022-11-14 01:02:14,950] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 053| loss: 0.701259| val_0_mae: 0.688155| 0:00:53s
(run_trial pid=42476) [2022-11-14 01:02:16,178] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 054| loss: 0.666418| val_0_mae: 0.683142| 0:00:54s
(run_trial pid=42476) [2022-11-14 01:02:17,413] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 055| loss: 0.680250| val_0_mae: 0.683969| 0:00:56s
(run_trial pid=42476) [2022-11-14 01:02:18,641] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 056| loss: 0.668339| val_0_mae: 0.695880| 0:00:57s
(run_trial pid=42476) [2022-11-14 01:02:19,877] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 057| loss: 0.692120| val_0_mae: 0.701301| 0:00:58s
(run_trial pid=42476) [2022-11-14 01:02:21,113] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 058| loss: 0.665579| val_0_mae: 0.690541| 0:00:59s
(run_trial pid=42476) [2022-11-14 01:02:22,363] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 059| loss: 0.656419| val_0_mae: 0.765069| 0:01:01s
(run_trial pid=42476) [2022-11-14 01:02:23,606] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 060| loss: 0.682641| val_0_mae: 0.802213| 0:01:02s
(run_trial pid=42476) [2022-11-14 01:02:24,863] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 061| loss: 0.643246| val_0_mae: 0.765911| 0:01:03s
(run_trial pid=42476) [2022-11-14 01:02:26,120] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 062| loss: 0.657356| val_0_mae: 0.743121| 0:01:04s
(run_trial pid=42476) [2022-11-14 01:02:27,378] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 063| loss: 0.704962| val_0_mae: 0.681874| 0:01:06s
(run_trial pid=42476) [2022-11-14 01:02:28,625] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 064| loss: 0.627800| val_0_mae: 0.684119| 0:01:07s
(run_trial pid=42476) [2022-11-14 01:02:29,886] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 065| loss: 0.635989| val_0_mae: 0.716252| 0:01:08s
(run_trial pid=42476) [2022-11-14 01:02:31,157] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 066| loss: 0.647488| val_0_mae: 0.694206| 0:01:09s
(run_trial pid=42476) [2022-11-14 01:02:32,441] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 067| loss: 0.667528| val_0_mae: 0.684332| 0:01:11s
(run_trial pid=42476) [2022-11-14 01:02:33,729] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 068| loss: 0.626634| val_0_mae: 0.667108| 0:01:12s
(run_trial pid=42476) [2022-11-14 01:02:35,006] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 069| loss: 0.627659| val_0_mae: 0.675814| 0:01:13s
(run_trial pid=42476) [2022-11-14 01:02:36,276] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 070| loss: 0.646840| val_0_mae: 0.710005| 0:01:15s
(run_trial pid=42476) [2022-11-14 01:02:37,545] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 071| loss: 0.630839| val_0_mae: 0.758983| 0:01:16s
(run_trial pid=42476) [2022-11-14 01:02:38,811] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 072| loss: 0.648888| val_0_mae: 0.736240| 0:01:17s
(run_trial pid=42476) [2022-11-14 01:02:40,079] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 073| loss: 0.635487| val_0_mae: 0.685271| 0:01:18s
(run_trial pid=42476) [2022-11-14 01:02:41,345] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 074| loss: 0.632486| val_0_mae: 0.720887| 0:01:20s
(run_trial pid=42476) [2022-11-14 01:02:42,615] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 075| loss: 0.646977| val_0_mae: 0.705984| 0:01:21s
(run_trial pid=42476) [2022-11-14 01:02:43,889] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 076| loss: 0.667855| val_0_mae: 0.678300| 0:01:22s
(run_trial pid=42476) [2022-11-14 01:02:45,162] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 077| loss: 0.642378| val_0_mae: 0.680050| 0:01:23s
(run_trial pid=42476) [2022-11-14 01:02:46,444] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 078| loss: 0.606339| val_0_mae: 0.718853| 0:01:25s
(run_trial pid=42476) [2022-11-14 01:02:47,736] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 079| loss: 0.610563| val_0_mae: 0.696663| 0:01:26s
(run_trial pid=42476) [2022-11-14 01:02:49,010] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 080| loss: 0.624385| val_0_mae: 0.738892| 0:01:27s
(run_trial pid=42476) [2022-11-14 01:02:50,293] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 081| loss: 0.650479| val_0_mae: 0.785787| 0:01:29s
(run_trial pid=42476) [2022-11-14 01:02:51,577] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 082| loss: 0.630244| val_0_mae: 0.725113| 0:01:30s
(run_trial pid=42476) [2022-11-14 01:02:52,875] [paddlets.models.common.callbacks.callbacks] [INFO] epoch 083| loss: 0.596762| val_0_mae: 0.722377| 0:01:31s
(run_trial pid=42476) [2022-11-14 01:02:52,876] [paddlets.models.common.callbacks.callbacks] [INFO]
(run_trial pid=42476) Early stopping occurred at epoch 83 with best_epoch = 38 and best_val_0_mae = 0.662641
(run_trial pid=42476) [2022-11-14 01:02:52,876] [paddlets.models.common.callbacks.callbacks] [INFO] Best weights from best epoch are automatically used!
(run_trial pid=42476) C:\Users\Administrator\anaconda3\lib\site-packages\ray\tune\trainable\session.py:229: DeprecationWarning: tune.report and tune.checkpoint_dir APIs are deprecated in Ray 2.0, and is replaced by ray.air.session. This will provide an easy-to-use API across Tune session and Data parallel worker sessions.The old APIs will be removed in the future.
(run_trial pid=42476) warnings.warn(

我现在在处理的问题是有很多个比较短的序列，需要对每个序列分别预测他的未来的值，但是我阅读paddlets的文档后似乎没有找到相关的api，请问我可以通过什么方法来解决这样的问题呢？
我的数据可以看成是假设有1000只股票的数据，每只股票有过去30个星期的数据，我需要预测未来10个星期每个股票的走势，请问paddlets是否提供了相应的api？或者我可以通过什么方法来解决这一问题？

系统版本：Ubuntu 22.04
电脑配置： i7-12700 + Nvidia RTX 3070
在conda环境中安装正常，import时报错

(pdts) ncg@ncg-pc:~$ pip list | grep paddlets
paddlets           1.0.1
(pdts) ncg@ncg-pc:~$ pip list | grep paddlepaddle
paddlepaddle-gpu   2.3.2.post111
(pdts) ncg@ncg-pc:~$ python
Python 3.8.13 (default, Oct 21 2022, 23:50:54) 
[GCC 11.2.0] :: Anaconda, Inc. on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> from paddlets import TSDataset
Error: Can not import avx core while this file exists: /home/ncg/anaconda3/envs/pdts/lib/python3.8/site-packages/paddle/fluid/core_avx.so
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/ncg/anaconda3/envs/pdts/lib/python3.8/site-packages/paddlets/__init__.py", line 5, in <module>
    from paddlets.pipeline import Pipeline
  File "/home/ncg/anaconda3/envs/pdts/lib/python3.8/site-packages/paddlets/pipeline/__init__.py", line 4, in <module>
    from paddlets.pipeline.pipeline import Pipeline
  File "/home/ncg/anaconda3/envs/pdts/lib/python3.8/site-packages/paddlets/pipeline/pipeline.py", line 16, in <module>
    from paddlets.utils.utils import get_tsdataset_max_len, split_dataset
  File "/home/ncg/anaconda3/envs/pdts/lib/python3.8/site-packages/paddlets/utils/__init__.py", line 8, in <module>
    from paddlets.utils.backtest import backtest
  File "/home/ncg/anaconda3/envs/pdts/lib/python3.8/site-packages/paddlets/utils/backtest.py", line 12, in <module>
    from paddlets.metrics import Metric, MSE
  File "/home/ncg/anaconda3/envs/pdts/lib/python3.8/site-packages/paddlets/metrics/__init__.py", line 7, in <module>
    from paddlets.metrics.metrics import (
  File "/home/ncg/anaconda3/envs/pdts/lib/python3.8/site-packages/paddlets/metrics/metrics.py", line 8, in <module>
    import paddle
  File "/home/ncg/anaconda3/envs/pdts/lib/python3.8/site-packages/paddle/__init__.py", line 25, in <module>
    from .framework import monkey_patch_variable
  File "/home/ncg/anaconda3/envs/pdts/lib/python3.8/site-packages/paddle/framework/__init__.py", line 17, in <module>
    from . import random  # noqa: F401
  File "/home/ncg/anaconda3/envs/pdts/lib/python3.8/site-packages/paddle/framework/random.py", line 16, in <module>
    import paddle.fluid as fluid
  File "/home/ncg/anaconda3/envs/pdts/lib/python3.8/site-packages/paddle/fluid/__init__.py", line 36, in <module>
    from . import framework
  File "/home/ncg/anaconda3/envs/pdts/lib/python3.8/site-packages/paddle/fluid/framework.py", line 37, in <module>
    from . import core
  File "/home/ncg/anaconda3/envs/pdts/lib/python3.8/site-packages/paddle/fluid/core.py", line 298, in <module>
    raise e
  File "/home/ncg/anaconda3/envs/pdts/lib/python3.8/site-packages/paddle/fluid/core.py", line 256, in <module>
    from . import core_avx
ImportError: /home/ncg/anaconda3/envs/pdts/lib/python3.8/site-packages/paddle/fluid/core_avx.so: undefined symbol: _dl_sym, version GLIBC_PRIVATE

NBEATSModel推理的时候指定的协变量名字和训练时候的协变量名字不一致，模型可以完成推理，这个还算可以理解，可以假设predict时候构造TSDataset填的observed_cov_cols列表是顺序相关的
但是，我主动改变observed_cov_cols协变量名字的顺序，预测结果没发生改变，这个似乎不太合理，参考如下代码

import numpy as np
import pandas as pd
import paddlets


if __name__ == "__main__":
    x = np.linspace(-np.pi, np.pi, 200)
    sinx = np.sin(x) * 4 + np.random.randn(200)

    df1 = pd.DataFrame(
        {
            "time_col1": pd.date_range("2022-01-01", periods=200, freq="1h"),
            "value_col1": sinx,
            "observed_col11": sinx + 1,
            "observed_col12": sinx + 2,
        }
    )
    ts1 = paddlets.TSDataset.load_from_dataframe(
        df1,
        time_col="time_col1",
        target_cols="value_col1",
        observed_cov_cols=["observed_col11", "observed_col12"],
    )

    # 训练模型
    train_ts, val_ts = ts1.split(0.8)
    from paddlets.models.forecasting import NBEATSModel

    nbeast = NBEATSModel(in_chunk_len=20, out_chunk_len=10, max_epochs=500)
    nbeast.fit(train_ts, val_ts)

    # 预测一，协变量的名字和训练时候协变量名字不同
    np.random.seed(1)
    df21 = pd.DataFrame(
        {
            "time_col2": pd.date_range("2022-01-01", periods=200, freq="1h"),
            "value_col2": sinx,
            "observed_col21": sinx - 20 * np.random.randn(200),
            "observed_col22": sinx + 30 * np.random.randn(200),
        }
    )
    ts21 = paddlets.TSDataset.load_from_dataframe(
        df21,
        time_col="time_col2",
        target_cols="value_col2",
        observed_cov_cols=["observed_col21", "observed_col22"],
        freq="1h",
    )
    predict_result_21 = nbeast.predict(ts21)

    # 预测二，重新生成TSDataset对象，但是指定observed_cov_cols的时候把名字对调
    ts22 = paddlets.TSDataset.load_from_dataframe(
        df21,
        time_col="time_col2",
        target_cols="value_col2",
        observed_cov_cols=["observed_col22", "observed_col21"],
        freq="1h",
    )
    predict_result_22 = nbeast.predict(ts22)

    # 打印两次预测结果
    print("predict 1:")
    print(predict_result_21)
    print("predict 2:")
    print(predict_result_22)
    print("预测用的ts的observed_cov_cols列名称和训练对不上，而且名称的顺序不影响预测结果，有点奇怪")

这边得到的结果如下：

predict 1:
                     value_col2
2022-01-09 08:00:00   29.420271
2022-01-09 09:00:00   30.838749
2022-01-09 10:00:00    4.161640
2022-01-09 11:00:00   24.282003
2022-01-09 12:00:00   30.760273
2022-01-09 13:00:00   -1.353035
2022-01-09 14:00:00   57.743870
2022-01-09 15:00:00   44.322231
2022-01-09 16:00:00   -2.120617
2022-01-09 17:00:00   16.197647
predict 2:
                     value_col2
2022-01-09 08:00:00   29.420271
2022-01-09 09:00:00   30.838749
2022-01-09 10:00:00    4.161640
2022-01-09 11:00:00   24.282003
2022-01-09 12:00:00   30.760273
2022-01-09 13:00:00   -1.353035
2022-01-09 14:00:00   57.743870
2022-01-09 15:00:00   44.322231
2022-01-09 16:00:00   -2.120617
2022-01-09 17:00:00   16.197647

看当前版本中DeepAR model没有在paddlets_default_search_space中，我自定义了一个search_space，

                            "rnn_type_or_module": choice(["GRU","LSTM"]),
                            "hidden_size": qrandint(32, 512, q=32),
                            "num_layers_recurrent": randint(1, 4),
                            "dropout": quniform(0, 1, 0.05),
                            "skip_chunk_len":0,
                            "regression_mode":choice(["mean", "sampling"]),
                            "batch_size":qrandint(8,128,q=8),
                            "max_epochs":qrandint(30,600,q=30),
                            "patience":qrandint(5,50,q=5),
                            "optimizer_params": {
                                "learning_rate": uniform(1e-4, 1e-2)
                            },
                            "seed":42
                        }

但是实际使用时提示

ray.exceptions.RayTaskError(ValueError): ray::ImplicitFunc.train() (pid=38826, ip=127.0.0.1, repr=run_trial)
  File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/ray/tune/trainable/trainable.py", line 355, in train
    raise skipped from exception_cause(skipped)
  File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/ray/tune/trainable/function_trainable.py", line 325, in entrypoint
    return self._trainable_func(
  File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/ray/tune/trainable/function_trainable.py", line 651, in _trainable_func
    output = fn()
  File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/paddlets/automl/optimize_runner.py", line 174, in run_trial
    score = fit_and_score(train_data=train_tsdataset,
  File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/paddlets/utils/validation.py", line 144, in fit_and_score
    score_dict, predicts = backtest(data,
  File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/paddlets/utils/backtest.py", line 150, in backtest
    score_dict = metric(real, predict)
  File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/paddlets/metrics/base.py", line 261, in __call__
    res_array = self._build_metrics_data(tsdataset_true, tsdataset_pred)
  File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/paddlets/metrics/base.py", line 54, in _build_metrics_data
    raise_if_not(
  File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/paddlets/logger/logger.py", line 135, in raise_if_not
    raise ValueError(message)
ValueError: In `normal` mode, only point forecasting data is supported!

请问这该如何修改呀

ml_base.py

第101行：

# b.modelname = "a"

应该是：

# b.modelname = "b"

我提交了PR。

用OneHot对tsdataset进行inverse_transform不起作用，无法复原被分解的列，无论drop是True还是False

`AutoTS自动寻参完成的模型如何保存精度最高的模型，用.save()方法暂不支持，用 best_param 返回的最佳参数去加载一个新的模型训练完保存吗

去运行时提示这些未能完成
Traceback (most recent call last):
File "/Users/susheng/PycharmProjects/paddleTest/AutoTS.py", line 124, in
autots_tcn.fit(ts_train, ts_val)
File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/paddlets/logger/logger.py", line 27, in wrapper
result = f(*args, **kwargs)
File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/paddlets/automl/autots.py", line 207, in fit
analysis = self._optimize_runner.optimize(self._estimator,
File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/paddlets/automl/optimize_runner.py", line 204, in optimize
return tune.run(run_trial, num_samples=n_trials, config=running_search_space,
File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/ray/tune/tune.py", line 771, in run
raise TuneError("Trials did not complete", incomplete_trials)
ray.tune.error.TuneError: ('Trials did not complete', [run_trial_732b269c, run_trial_88119474, run_trial_7c009fba, run_trial_94cc868a, run_trial_ab0a0170, run_trial_a7e53298, run_trial_afbec2c2, run_trial_be90113e])

然后我去看快ray-result对应trail的报错文档，发现他们基本显示的
Failure # 1 (occurred at 2022-11-17_19-39-14)
�[36mray::ImplicitFunc.train()�[39m (pid=30278, ip=127.0.0.1, repr=run_trial)
File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/ray/tune/trainable/trainable.py", line 355, in train
raise skipped from exception_cause(skipped)
File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/ray/tune/trainable/function_trainable.py", line 325, in entrypoint
return self._trainable_func(
File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/ray/tune/trainable/function_trainable.py", line 651, in _trainable_func
output = fn()
File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/paddlets/automl/optimize_runner.py", line 174, in run_trial
score = fit_and_score(train_data=train_tsdataset,
File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/paddlets/utils/validation.py", line 127, in fit_and_score
estimator.fit(train_data, valid_data)
File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/paddlets/models/forecasting/dl/paddle_base_impl.py", line 346, in fit
self._fit(train_dataloader, valid_dataloaders)
File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/paddlets/models/forecasting/dl/paddle_base_impl.py", line 363, in _fit
self._network = self._init_network()
File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/paddlets/models/forecasting/dl/tcn.py", line 341, in _init_network
return _TCNModule(
File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/paddlets/models/forecasting/dl/tcn.py", line 126, in init
raise_if_not(
File "/Users/susheng/opt/anaconda3/envs/paddle/lib/python3.8/site-packages/paddlets/logger/logger.py", line 135, in raise_if_not
raise ValueError(message)
ValueError: The valid range of kernel_size is (1, in_chunk_len], got kernel_size:1 <= 1 or kernel_size:1 > in_chunk_len:24.

看起来像kernel-size不合规导致的，请问这该如何修改呢？

paddlets\tests\transform\test_normalization.py

第50行：

self.assertTrue(inverse1.get_known_cov().data.astype('int').equals(input1.get_known_cov().data))

inverse1.get_known_cov().data.astype('int') 与 input1.get_known_cov().data 永远不相等，

因为前者的dtype是int32，而后者是int64。如果修改成：

self.assertTrue(inverse1.get_known_cov().data.astype('int').equals(input1.get_known_cov().data.astype('int')))

则可以顺利通过。

该单元测试中test_MinMax与test_StandardScaler这行代码出现了8次，都存在这个情况。

Python3.8 64bit

另外我发现该单元测试中test_with_sklearn也有类似的相等判断，但是两侧已经都加了.astype('int')，所以可以顺利通过。

ubuntu 20.04
paddlets 1.0.1
##########################################################################################

ts2vec_params = {"segment_size": 200,
"repr_dims": 320,
"batch_size": 32,
"sampling_stride": 200,
"max_epochs": 20}
model = ReprForecasting(in_chunk_len=200,
out_chunk_len=24,
sampling_stride=1,
repr_model=TS2Vec,
repr_model_params=ts2vec_params)
model.fit(train_data)
model.save("/weights/models1")
model=load("/weights/models1")

发生异常: ValueError
path must be a file path, not a directory: /home/kylin/pythondevelop/paddle_tutorial/ts/trade/weights/models1

想问下PaddleTS具体哪些模型是支持协变量的，从文档里面没找到很好的说明
然后用MLPRegressor模型测试了下，训练用的TSDataset是包含observed_cov_cols，但是预测的时候传递的TSDataset不包含observed_cov_cols，还是能正确推理，例如下面这段代码：

import numpy as np
import pandas as pd
import paddlets

if __name__ == "__main__":
    x = np.linspace(-np.pi, np.pi, 200)
    sinx = np.sin(x) * 4 + np.random.randn(200)

    df1 = pd.DataFrame(
        {
            "time_col1": pd.date_range("2022-01-01", periods=200, freq="1h"),
            "value_col1": sinx,
            "observed_col1": sinx + 1,
        }
    )
    ts1 = paddlets.TSDataset.load_from_dataframe(
        df1,
        time_col="time_col1",
        target_cols="value_col1",
        observed_cov_cols=["observed_col1"],
    )

    # 训练模型
    train_ts, val_ts = ts1.split(0.8)
    from paddlets.models.forecasting import MLPRegressor

    mlp = MLPRegressor(in_chunk_len=20, out_chunk_len=10, max_epochs=500)
    mlp.fit(train_ts, val_ts)

    # 预测模型
    df2 = pd.DataFrame(
        {
            "time_col2": pd.date_range("2022-01-01", periods=200, freq="1h"),
            "value_col2": sinx,
        }
    )
    ts2 = paddlets.TSDataset.load_from_dataframe(
        df2, time_col="time_col2", target_cols="value_col2", freq="1h"
    )
    predict_result_2 = mlp.predict(ts2)
    print("predict results: ", predict_result_2)

首先感谢分享！我目前主要研究异常检测学习。
在如下链接：
https://paddlets.readthedocs.io/zh_CN/latest/source/modules/models/anomaly.html
其中 5. 模型预测和评估 的指标计算，未见 res 从哪里来？应该是 pred_label 的吧？
pred_label = model.predict(test_data_scaled) lable_name = pred_label.target.data.columns[0] f1 = F1()(test_tsdata, **res**) precision = Precision()(test_tsdata, **res**) recall = Recall()(test_tsdata, **res**)

代码如下：

ix_dataset2 = TSDataset.load_from_dataframe(
    data3,
    time_col='time',
    target_cols='ix',
    freq='1d',
    known_cov_cols=['work_day','holiday']
)

from paddlets.models.forecasting import MLPRegressor,LSTNetRegressor,TCNRegressor
md = TCNRegressor(
    in_chunk_len = 7,
    out_chunk_len = 1,
    max_epochs=100,
)

train_dataset, val_test_dataset = ix_dataset2.split(0.7)
val_dataset, test_dataset = val_test_dataset.split(0.5)
# train_dataset.plot(add_data=[val_dataset,test_dataset], labels=['Val', 'Test'])

md.fit(train_dataset, val_dataset)

from paddlets.utils import backtest
score, preds_data= backtest(
    data=test_dataset,
    model=md,
    return_predicts = True)
test_dataset.plot(add_data=preds_data,labels="backtest")

结果如下：

预测趋势相当不错，但是存在明显偏移，请问这个是什么原因？求解答

1.千呼万唤始出来,终于有了飞桨框架的时间序列预测了.
2.我进入时间序列预测领域,都是由于获得AAAI2021最佳论文奖的那篇论文引起我的兴趣的.AAAI 2021最佳论文：比Transformer更有效的长时间序列预测.看到这个标题,竟然比Transformer更牛逼,万分吃惊.请问PaddleTS大佬,几时复现这篇论文呢?或者是复现2022年比较牛逼的时间序列预测的论文.如果我记得不错的话,AAAI2021最佳论文,是有pytorch的复现版本的,就在github.其实多数论文都有pytorch的复现版本,毕竟科研者写论文,超过80%以上的论文就是使用pytorch框架.使用飞桨框架复现,可以参考参考.
3.看到大佬们对于未来的工作的概括,小弟最感兴趣的就是概率模型,希望早日复现该领域最牛逼的论文.我听说有人使用概率模型炒股,发了大财,不知道真假,哈哈哈哈.
4.图像分类的数据集,有一个很大的imagenet数据集,请问时间序列预测,有什么很大的数据集吗?目前PaddleTS提供了很多模型,请问你们认为哪个模型效果最好呢?哪个模型适用于短时间序列预测,哪个模型适用于长时间序列预测?哪个模型适用于小数据集,哪个模型适用于大数据集?
5.我是看了飞桨视觉的负责人的一篇微信朋友圈,才知道PaddleTS这个仓库.朋友圈中介绍,时间序列预测可以用于股价预测,请问你们认同时间序列预测可以用于股价预测这种观点吗?祝你们炒股赚大钱,发大财.

paddlets\transform\base.py

第219行：

return self.inverse_transform_one(dataset)

没有传入 inplace 参数，这个参数不起作用。

PR #286

W0915 12:02:58.731670 1250 gpu_resources.cc:61] Please NOTE: device: 0, GPU Compute Capability: 8.6, Driver API Version: 11.7, Runtime API Version: 11.7
W0915 12:02:58.744315 1250 gpu_resources.cc:91] device: 0, cuDNN Version: 8.5.
Traceback (most recent call last):
File "test.py", line 34, in
model.fit(train_dataset, val_dataset)
File "/usr/local/lib/python3.8/dist-packages/paddlets/models/dl/paddlepaddle/paddle_base_impl.py", line 321, in fit
self._fit(train_dataloader, valid_dataloaders)
File "/usr/local/lib/python3.8/dist-packages/paddlets/models/dl/paddlepaddle/paddle_base_impl.py", line 347, in _fit
self._train_epoch(train_dataloader)
File "/usr/local/lib/python3.8/dist-packages/paddlets/models/dl/paddlepaddle/paddle_base_impl.py", line 415, in _train_epoch
batch_logs = self._train_batch(X, y)
File "/usr/local/lib/python3.8/dist-packages/paddlets/models/dl/paddlepaddle/paddle_base_impl.py", line 434, in _train_batch
output = self._network(X)
File "/usr/local/lib/python3.8/dist-packages/paddle/fluid/dygraph/layers.py", line 930, in call
return self._dygraph_call_func(*inputs, **kwargs)
File "/usr/local/lib/python3.8/dist-packages/paddle/fluid/dygraph/layers.py", line 915, in _dygraph_call_func
outputs = self.forward(*inputs, **kwargs)
File "/usr/local/lib/python3.8/dist-packages/paddlets/models/dl/paddlepaddle/transformer.py", line 196, in forward
out = self._transformer(src, tgt)
File "/usr/local/lib/python3.8/dist-packages/paddle/fluid/dygraph/layers.py", line 930, in call
return self._dygraph_call_func(*inputs, **kwargs)
File "/usr/local/lib/python3.8/dist-packages/paddle/fluid/dygraph/layers.py", line 915, in _dygraph_call_func
outputs = self.forward(*inputs, **kwargs)
File "/usr/local/lib/python3.8/dist-packages/paddle/nn/layer/transformer.py", line 1628, in forward
memory = self.encoder(src, src_mask=src_mask)
File "/usr/local/lib/python3.8/dist-packages/paddle/fluid/dygraph/layers.py", line 930, in call
return self._dygraph_call_func(*inputs, **kwargs)
File "/usr/local/lib/python3.8/dist-packages/paddle/fluid/dygraph/layers.py", line 915, in _dygraph_call_func
outputs = self.forward(*inputs, **kwargs)
File "/usr/local/lib/python3.8/dist-packages/paddle/nn/layer/transformer.py", line 971, in forward
src_mask = _convert_attention_mask(src_mask, src.dtype,
File "/usr/local/lib/python3.8/dist-packages/paddle/nn/layer/transformer.py", line 108, in _convert_attention_mask
mha_meta = _prepare_mha_meta(attn_mask, enable_cudnn)
File "/usr/local/lib/python3.8/dist-packages/paddle/nn/layer/transformer.py", line 134, in _prepare_mha_meta
assert attn_mask is not None,
AssertionError: The attention mask should be given for MultiHeadAttention when enable_cudnn=True. But received attn_mask = None

from paddlets.utils import backtest q_loss, quantiles = backtest(data=ts_test_scaled, model=deepar, start="2013-11-01 00:00:00", metric=QuantileLoss([0.1, 0.5, 0.9]), predict_window=24, stride=24, return_predicts=True )

MergeError: No common columns to perform merge on. Merge options: left_on=None, right_on=None, left_index=False, right_index=False

ImportError: cannot import name 'hparams' from 'tensorboardX.summary' (/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/tensorboardX/summary.py)

`import pandas as pd
import os

import shutil
from paddlets.models.forecasting import MLPRegressor
from paddlets.automl.autots import AutoTS

通过pandas读取数据并显示前三条，与上图一一对应

csv_path = "tu_share_data_day/688388.SH.csv"

tu_share_data_day = os.listdir("tu_share_data_day")
for one in tu_share_data_day:
csv_path = os.path.join("tu_share_data_day", one)

dam_data = pd.read_csv(csv_path).values.tolist()
if len(dam_data) < 100:
    continue
dam_data.reverse()
for i in range(len(dam_data) - 1):
    dam_data[i].append(dam_data[i + 1][4])
    dam_data[i].append(dam_data[i + 1][5])

    dam_data[i].append(dam_data[i + 1][6])
    dam_data[i].append(int(i + 1))
# print(dam_data)
dam_data = pd.DataFrame(dam_data[:-1],
                        columns=['ts_code', 'trade_date', 'open', 'high', 'low', 'close',
                                 'pre_close', 'change', 'pct_chg', 'vol', 'amount', 'next_high',
                                 'next_low',
                                 'next_close', 'index_new'])
from paddlets.datasets import TSDataset

# 构建数据集
dataset = TSDataset.load_from_dataframe(
    dam_data,  # pd.DataFrame
    time_col="index_new",  # 索引
    target_cols=['next_high', 'next_low',
                 'next_close'],  # 需要预测的结果
    observed_cov_cols=['open', 'high', 'low', 'close',
                       'pre_close', 'change', 'pct_chg', 'vol', 'amount']  # 观测协变量
)
# 划分训练集和验证集
train_dataset, val_dataset = dataset.split(0.8)  # 8:2

autots_model = AutoTS(MLPRegressor, 96, 2)
autots_model.fit(train_dataset, val_dataset)`

import paddlets
from paddlets.datasets.repository import get_dataset, dataset_list

print(paddlets.version)
0.1.0

print(f"built-in datasets:{dataset_list()}")
built-in datasets:['UNI_WTH', 'ETTh1', 'ETTm1', 'ECL', 'WTH']

dataset = get_dataset('WTH')
执行到 dataset = get_dataset('WTH') 这行时提示
[2022-08-11 14:07:44,459] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!
[2022-08-11 14:07:44,487] [paddlets] [ERROR] ValueError: attr: shape doesn't exist!

加载数据集

ts_data = load_datasets.load_datasets_anomaly(time_col='Date',
feature_cols=["L#SVR4_VAR830_A","L#SVR4_VAR833_A","L#SVR4_VAR476_A","L#SVR4_VAR350_A","L#SVR4_VAR340_A","L#SVR4_VAR342_A","L#SVR4_VAR344_A"]
)

可视化数据集

from paddlets.utils.utils import plot_anoms
plot_anoms(origin_data=ts_data, feature_name="L#SVR4_VAR830_A")

报错如下：

AttributeError Traceback (most recent call last)
Cell In [16], line 4
2 matplotlib.use('TkAgg') #加上，不然显示图表有问题
3 import matplotlib.pyplot as plt
----> 4 ts_data.plot(add_data=["L#SVR4_VAR830_A","L#SVR4_VAR833_A","L#SVR4_VAR476_A","L#SVR4_VAR350_A","L#SVR4_VAR340_A","L#SVR4_VAR342_A","L#SVR4_VAR344_A"])
5 plt.show()
6 # from paddlets.utils.utils import plot_anoms
7 #
8 # plot_anoms(origin_data=ts_data, feature_name="L#SVR4_VAR830_A")

File E:\Paddle-release-2.2\PaddleTS\paddlets\datasets\tsdataset.py:1689, in TSDataset.plot(self, columns, add_data, labels, low_quantile, high_quantile, central_quantile, **kwargs)
1656 def plot(self,
1657 columns:Union[List[str], str] = None,
1658 add_data:Union[List["TSDataset"], "TSDataset"] = None,
(...)
1662 central_quantile:float = 0.5,
1663 **kwargs) -> "pyplot":
1664 """
1665 plot function, a wrapper for Dataframe.plot()
1666
(...)
1687
1688 """
-> 1689 quantile_cols = self._get_quantile_cols_origin_names()
1690 if not columns:
1691 if len(quantile_cols) == 0:

File E:\Paddle-release-2.2\PaddleTS\paddlets\datasets\tsdataset.py:1849, in TSDataset._get_quantile_cols_origin_names(self)
1841 """
1842 Get quantile cols origin names
1843
(...)
1846
1847 """
1848 origin_columns = []
-> 1849 for name in self.target.columns:
1850 tmp = name.split("@quantile")
1851 if tmp[0] not in origin_columns and len(tmp) > 1:

AttributeError: 'NoneType' object has no attribute 'columns'

请问一下表征模型是否支持多变量，在dataset中设置了known_cov_cols，然后使用CoST表征模型，但似乎没起作用，请问表征模型是否支持进行多变量训练，如果没有的话，如何才能在表征模型中使用多变量呢？

修改metric参数为mse，但实际训练评估使用的仍是mae。
metric (str) – metric的名字。metric 会被用于计算在 validation 数据集上的 loss，并反馈给超参优化算法。

加载已保存的模型

def load_model(self,
load_pdmodel: str = '../model/save_model.pdmodel',
load_pdiparams: str = '../model/save_model.pdiparams'
):
config = paddle_infer.Config(load_pdmodel, load_pdiparams)
self.predictor = paddle_infer.create_predictor(config)
input_names = self.predictor.get_input_names()
print(f"input_name: f{input_names}")
import json
with open("../model/save_model_model_meta") as f:
self.json_data = json.load(f)
print(self.json_data)

预测推理

def predict_ts(self,
               pre_tsdataset: TSDataset,
               cols,
               recursive: bool = False):
    val_dataset = self.data_transform(tsdataset=pre_tsdataset,
                                      cols=cols)
    from paddlets.utils.utils import build_ts_infer_input
    input_data = build_ts_infer_input(val_dataset, "../model/save_model_model_meta")

    for key, value in self.json_data['input_data'].items():
        input_handle1 = self.predictor.get_input_handle(key)
        # set batch_size=1
        value[0] = 1
        input_handle1.reshape(value)
        input_handle1.copy_from_cpu(input_data[key])

    self.predictor.run()
    output_names = self.predictor.get_output_names()
    output_handle = self.predictor.get_output_handle(output_names[0])
    output_data = output_handle.copy_to_cpu()
    print(output_data)

##############################################################################################

调用过程：

# 模型加载

fit_lstnet.load_model(load_pdmodel='../model/LSTNet.pdmodel',
load_pdiparams='../model/LSTNet.pdiparams')

# 预测推理

fit_lstnet.predict_ts(pre_tsdataset=fit_lstnet.val_dataset, cols=target_cols)

执行load_model保存模型成功
执行predict_ts函数时build_ts_infer_input报错，报错信息如下：
e[1me[35m--- Running analysis [ir_graph_build_pass]e[0m
e[1me[35m--- Running analysis [ir_graph_clean_pass]e[0m
e[1me[35m--- Running analysis [ir_analysis_pass]e[0m
e[32m--- Running IR pass [simplify_with_basic_ops_pass]e[0m
e[32m--- Running IR pass [layer_norm_fuse_pass]e[0m
e[37m--- Fused 0 subgraphs into layer_norm op.e[0m
e[32m--- Running IR pass [attention_lstm_fuse_pass]e[0m
e[32m--- Running IR pass [seqconv_eltadd_relu_fuse_pass]e[0m
e[32m--- Running IR pass [seqpool_cvm_concat_fuse_pass]e[0m
e[32m--- Running IR pass [mul_lstm_fuse_pass]e[0m
e[32m--- Running IR pass [fc_gru_fuse_pass]e[0m
e[37m--- fused 0 pairs of fc gru patternse[0m
e[32m--- Running IR pass [mul_gru_fuse_pass]e[0m
e[32m--- Running IR pass [seq_concat_fc_fuse_pass]e[0m
e[32m--- Running IR pass [gpu_cpu_squeeze2_matmul_fuse_pass]e[0m
e[32m--- Running IR pass [gpu_cpu_reshape2_matmul_fuse_pass]e[0m
e[32m--- Running IR pass [gpu_cpu_flatten2_matmul_fuse_pass]e[0m
e[32m--- Running IR pass [matmul_v2_scale_fuse_pass]e[0m
e[32m--- Running IR pass [gpu_cpu_map_matmul_v2_to_mul_pass]e[0m
I1230 08:59:19.689038 23564 fuse_pass_base.cc:57] --- detected 2 subgraphs
e[32m--- Running IR pass [gpu_cpu_map_matmul_v2_to_matmul_pass]e[0m
e[32m--- Running IR pass [matmul_scale_fuse_pass]e[0m
e[32m--- Running IR pass [gpu_cpu_map_matmul_to_mul_pass]e[0m
e[32m--- Running IR pass [fc_fuse_pass]e[0m
I1230 08:59:19.691967 23564 fuse_pass_base.cc:57] --- detected 2 subgraphs
e[32m--- Running IR pass [repeated_fc_relu_fuse_pass]e[0m
e[32m--- Running IR pass [squared_mat_sub_fuse_pass]e[0m
e[32m--- Running IR pass [conv_bn_fuse_pass]e[0m
e[32m--- Running IR pass [conv_eltwiseadd_bn_fuse_pass]e[0m
e[32m--- Running IR pass [conv_transpose_bn_fuse_pass]e[0m
e[32m--- Running IR pass [conv_transpose_eltwiseadd_bn_fuse_pass]e[0m
e[32m--- Running IR pass [is_test_pass]e[0m
e[32m--- Running IR pass [runtime_context_cache_pass]e[0m
e[1me[35m--- Running analysis [ir_params_sync_among_devices_pass]e[0m
e[1me[35m--- Running analysis [adjust_cudnn_workspace_size_pass]e[0m
e[1me[35m--- Running analysis [inference_op_replace_pass]e[0m
e[1me[35m--- Running analysis [ir_graph_to_program_pass]e[0m
I1230 08:59:19.704661 23564 analysis_predictor.cc:1035] ======= optimize end =======
I1230 08:59:19.704661 23564 naive_executor.cc:102] --- skip [feed], feed -> past_target
I1230 08:59:19.705637 23564 naive_executor.cc:102] --- skip [tmp_0], fetch -> fetch
input_name: f['past_target']
{'model_type': 'forecasting', 'ancestor_classname_set': ['LSTNetRegressor', 'PaddleBaseModelImpl', 'PaddleBaseModel', 'BaseModel', 'Trainable', 'ABC', 'object'], 'modulename': 'paddlets.models.forecasting.dl.lstnet', 'size': {'in_chunk_len': 3600, 'out_chunk_len': 3600, 'skip_chunk_len': 0}, 'input_data': {'past_target': [None, 3600, 1]}}
Traceback (most recent call last):
File "E:/Paddle-release-2.2/PaddleTS/examples_self/fit_forecasting_model.py", line 410, in
fit_lstnet.predict_ts(pre_tsdataset=fit_lstnet.val_dataset, cols=target_cols)
File "E:/Paddle-release-2.2/PaddleTS/examples_self/fit_forecasting_model.py", line 284, in predict_ts
input_data = build_ts_infer_input(val_dataset, "../model/LSTNet_model_meta")
File "E:\Paddle-release-2.2\PaddleTS\paddlets\utils\utils.py", line 439, in build_ts_infer_input
sample = next(iter(dataloader))
File "F:\CondaData\envs\paddlets\lib\site-packages\paddle\fluid\dataloader\dataloader_iter.py", line 298, in next
six.reraise(*sys.exc_info())
File "F:\CondaData\envs\paddlets\lib\site-packages\six.py", line 719, in reraise
raise value
File "F:\CondaData\envs\paddlets\lib\site-packages\paddle\fluid\dataloader\dataloader_iter.py", line 272, in next
data = self._reader.read_next_var_list()
StopIteration

之前使用过飞浆的cv和nlp，偶然看到发布的paddlets，可实现对时间序列的预测，刚好未来有这方面的需求，马上来体验下！
运行mlp.fit(train_dataset, val_dataset)报错：
ValueError: only the following paddle versions are supported: {'2.2.0', '2.3.0'}, current paddle version: 2.3.1

综述

本提案旨在建议将NLinear时序预测模型加入PaddleTS模型库。

👉NLinear原作论文
👉基于PaddleTS的NLinear实现

本实现立足于原作设计，亦做出了以下改进：

⭐支持已知协变量与观测协变量。
⭐Linear结构可以按需增加隐层。

本提案将NLinear与结构类似的MLP以及主流的NBEATS进行了效果比较（见有效性验证一节）。
实验证明NLinear能以极低的模型复杂度实现较优异的效果。本实现做出的改进则进一步提升了NLinear的效果和泛用性。

NLinear模型介绍

NLinear是一个非常简单的线性模型，仅将历史输入通过一个线性层，就得到预测输出。

为了解决时序数据的分布漂移问题，NLinear又对输入输出做了一次非常简单的Normalization。以历史输入的最后一个数作为尾数。历史输入先减去尾数，然后通过线性层，得到的输出再加回尾数，作为模型的预测输出。

本实现做出的改进

支持已知协变量与观测协变量

原作设计虽然简单高效，但是并不能利用TSDataset中包含的协变量信息，对于协变量至关重要的任务并不适用，为了解决这个问题，本实现做了相应改进。在构建输入张量时，将PAST_TARGET、KNOWN_COV与OBSERVED_COV一维展平，然后横向拼接作为输入，得到一维输出后，再重新折叠成所需要的张量形状。

# concat backcast, known_cov, observed_cov if any
feature = [backcast.reshape((batch_size, 1, -1))]
if known_cov is not None:
    feature.append(known_cov.reshape((batch_size, 1, -1)))
if observed_cov is not None:
    feature.append(observed_cov.reshape((batch_size, 1, -1)))
out = paddle.concat(x=feature, axis=2)

# forward
out = self._nn(out)
out = out.reshape([batch_size, -1, self._target_dim])

Linear结构可以按需增加隐层

协变量增加了输入张量的复杂性，单线性层可能无法充分学习特征，导致欠拟合。为了解决这个问题，本实现也做了相应改进。增加了与MLP类似的hidden_config属性，允许开发者按需增加隐层。

# Unlike MLP, `hidden_config` is empty by default, so there can be no hidden layer.
layers = []
dims = [in_chunk_len_multi] + hidden_config + [out_chunk_len_multi]
for i in range(len(dims) - 2):
    layers.append(paddle.nn.Linear(dims[i], dims[i + 1]))
    if use_bn:
        layers.append(paddle.nn.BatchNorm1D(1))
    layers.append(paddle.nn.ReLU())
layers.append(paddle.nn.Linear(dims[-2], dims[-1]))
self._nn = paddle.nn.Sequential(*layers)

有效性验证

以下是NLinear与其它模型在主流数据集的实验结果，考虑到MLP并不支持协变量，且数据集中协变量不一定对预测有正向作用，所以每个数据集都做了有协变量与无协变量两遍测试。

隐层的有效性

从实验结果可知，虽然单线性层的NLinear效果不错，但是有隐层的NLinear更优，尤其当协变量多导致输入复杂时，效果尤为明显。因此可以认为本实现的可增加隐层改进是有效的。

Normalization的有效性

有隐层的NLinear换种角度看，可以被认为是一种增加了Normalization的MLP。从实验结果可知，当输入相同（都没有协变量），且网络结构一致（都有一个相同的隐层）时，NLinear均优于MLP。因此可以认为原作的Normalization特性是一个有效的特性。

协变量的有效性

并不是所有数据集的协变量都对预测有正向作用。通过与同样支持协变量的NBEATS对比可知，当协变量有助益时，NLinear模型可以从协变量中学习有效特征，实现比无协变量时更好的效果。因此可以认为本实现的支持协变量改进是有效的。（而MLP完全不支持协变量，对此类任务无适用性。）

结论

总的来说，NLinear简单好用，而本实现的NLinear模型在原作设计的基础上，进一步提升了效果和泛用性。NLinear的模型复杂程度并不比MLP高，却能实现比某些复杂模型更好的效果。模型调参简单，训练快，可以迅速地得出结论，适合作为各种任务的入手模型。

另外从PaddleTS自身发展的角度来看，NLinear无疑是对现有模型库的有益补充。考虑到开源社区内已有顶流时序建模库（5K+⭐）支持了该模型，PaddleTS吸纳NLinear亦有利于提高自身竞争力。

所以，建议PaddleTS加入NLinear时序预测模型。

Traceback (most recent call last):
File "D:\ProgramData\Anaconda3\lib\site-packages\paddlets\models\forecasting\dl\paddle_base.py", line 233, in save
pickle.dump(self, f)
TypeError: cannot pickle 'Tensor' object

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "D:\work\py\dossen-alg-api\paddle_tft.py", line 220, in
main()
File "D:\work\py\dossen-alg-api\paddle_tft.py", line 198, in main
model = train(train_dataset, valid_dataset)
File "D:\work\py\dossen-alg-api\paddle_tft.py", line 162, in train
model.save(RNN_MODEL_PATH)
File "D:\ProgramData\Anaconda3\lib\site-packages\paddlets\models\forecasting\dl\paddle_base.py", line 235, in save
raise_log(ValueError("error occurred while saving %s, err: %s" % (abs_model_path, str(e))))
File "D:\ProgramData\Anaconda3\lib\site-packages\paddlets\logger\logger.py", line 112, in raise_log
raise exception
ValueError: error occurred while saving D:\work\py\dossen-alg-api\model\tft\tft_model, err: cannot pickle 'Tensor' object

我按照官方文档https://paddlets.readthedocs.io/zh_CN/latest/source/get_started/run_on_gpu.html运行GPU实验代码：

import numpy as np

from paddlets.datasets.repository import get_dataset
from paddlets.transform.normalization import StandardScaler
from paddlets.models.forecasting import MLPRegressor

np.random.seed(2022)

# prepare data
tsdataset = get_dataset("WTH")
ts_train, ts_val_test = ts.split("2012-03-31 23:00:00")
ts_val, ts_test = ts_val_test.split("2013-02-28 23:00:00")

# transform
scaler = StandardScaler()
scaler.fit(ts_train)
ts_train_scaled = scaler.transform(ts_train)
ts_val_scaled = scaler.transform(ts_val)
ts_test_scaled = scaler.transform(ts_test)
ts_val_test_scaled = scaler.transform(ts_val_test)

# model
model = MLPRegressor(
     in_chunk_len=7 * 24,
     out_chunk_len=24,
     skip_chunk_len=0,
     sampling_stride=24,
     eval_metrics=["mse", "mae"],
     batch_size=32,
     max_epochs=1000,
     patience=100,
     use_bn=True,
     seed=2022
)

model.fit(ts_train_scaled, ts_val_scaled)

predicted_tsdataset = model.predict(ts_val_test_scaled)

print(predicted_tsdataset)

#                      WetBulbCelsius
# 2014-01-01 00:00:00       -0.124221
# 2014-01-01 01:00:00       -0.184970
# 2014-01-01 02:00:00       -0.398122
# 2014-01-01 03:00:00       -0.500016
# 2014-01-01 04:00:00       -0.350443
# 2014-01-01 05:00:00       -0.580986
# 2014-01-01 06:00:00       -0.482264
# 2014-01-01 07:00:00       -0.413248
# 2014-01-01 08:00:00       -0.451982
# 2014-01-01 09:00:00       -0.471430
# 2014-01-01 10:00:00       -0.427212
# 2014-01-01 11:00:00       -0.264509
# 2014-01-01 12:00:00       -0.308266
# 2014-01-01 13:00:00       -0.386270
# 2014-01-01 14:00:00       -0.261341
# 2014-01-01 15:00:00       -0.492441
# 2014-01-01 16:00:00       -0.497322
# 2014-01-01 17:00:00       -0.628926
# 2014-01-01 18:00:00       -0.528971
# 2014-01-01 19:00:00       -0.588881
# 2014-01-01 20:00:00       -0.860580
# 2014-01-01 21:00:00       -0.742121
# 2014-01-01 22:00:00       -0.819053
# 2014-01-01 23:00:00       -0.875322

遇到了一些Bug：

Traceback (most recent call last):
  File "old.py", line 4, in <module>
    from paddlets.transform.normalization import StandardScaler
ModuleNotFoundError: No module named 'paddlets.transform.normalization'

Traceback (most recent call last):
  File "old.py", line 11, in <module>
    ts_train, ts_val_test = ts.split("2012-03-31 23:00:00")
NameError: name 'ts' is not defined

我已经在PR中解决了上述Bug。https://github.com/PaddlePaddle/PaddleTS/pull/230

执行代码

官方示例

from paddlets.datasets.repository import get_dataset
tsdataset = get_dataset("UNI_WTH")
from paddlets.models.forecasting import MLPRegressor
from paddlets.automl.autots import AutoTS
autots_model = AutoTS(MLPRegressor, 96, 2)
autots_model.fit(tsdataset)

运行得到输出如下

加粗部分为报错，然后程序不终止，一直不断打印==Status==之后的内容，Number of trials也不变，这个是什么原因呢？

C:\ProgramData\Anaconda3\lib\site-packages\paddlets\automl\searcher.py:4: DeprecationWarning: The module ray.tune.suggest has been moved to ray.tune.search and the old location will be deprecated soon. Please adjust your imports to point to the new location. Example: Do a global search and replace ray.tune.suggest with ray.tune.search.
from ray.tune.suggest import BasicVariantGenerator
C:\ProgramData\Anaconda3\lib\site-packages\paddlets\automl\searcher.py:5: DeprecationWarning: The module ray.tune.suggest.optuna has been moved to ray.tune.search.optuna and the old location will be deprecated soon. Please adjust your imports to point to the new location. Example: Do a global search and replace ray.tune.suggest.optuna with ray.tune.search.optuna.
from ray.tune.suggest.optuna import OptunaSearch
C:\ProgramData\Anaconda3\lib\site-packages\paddlets\automl\searcher.py:6: DeprecationWarning: The module ray.tune.suggest.flaml has been moved to ray.tune.search.flaml and the old location will be deprecated soon. Please adjust your imports to point to the new location. Example: Do a global search and replace ray.tune.suggest.flaml with ray.tune.search.flaml.
from ray.tune.suggest.flaml import CFO
C:\ProgramData\Anaconda3\lib\site-packages\paddlets\automl\searcher.py:8: DeprecationWarning: The module ray.tune.suggest.bohb has been moved to ray.tune.search.bohb and the old location will be deprecated soon. Please adjust your imports to point to the new location. Example: Do a global search and replace ray.tune.suggest.bohb with ray.tune.search.bohb.
from ray.tune.suggest.bohb import TuneBOHB
C:\ProgramData\Anaconda3\lib\site-packages\paddlets\automl\search_space_configer.py:8: DeprecationWarning: The module ray.tune.sample has been moved to ray.tune.search.sample and the old location will be deprecated soon. Please adjust your imports to point to the new location. Example: Do a global search and replace ray.tune.sample with ray.tune.search.sample.
from ray.tune.sample import Float, Integer, Categorical
2022-12-27 14:21:00,844 INFO worker.py:1538 -- Started a local Ray instance.
C:\ProgramData\Anaconda3\lib\site-packages\ray\tune\search\optuna\optuna_search.py:694: FutureWarning: IntUniformDistribution has been deprecated in v3.0.0. This feature will be removed in v6.0.0. See https://github.com/optuna/optuna/releases/tag/v3.0.0. Use :class:~optuna.distributions.IntDistribution instead.
return ot.distributions.IntUniformDistribution(
C:\ProgramData\Anaconda3\lib\site-packages\ray\tune\search\optuna\optuna_search.py:682: FutureWarning: UniformDistribution has been deprecated in v3.0.0. This feature will be removed in v6.0.0. See https://github.com/optuna/optuna/releases/tag/v3.0.0. Use :class:~optuna.distributions.FloatDistribution instead.
return ot.distributions.UniformDistribution(
[I 2022-12-27 14:21:02,351] A new study created in memory with name: optuna
C:\ProgramData\Anaconda3\lib\site-packages\optuna\distributions.py:766: FutureWarning: IntUniformDistribution(high=128, low=8, step=8) is deprecated and internally converted to IntDistribution(high=128, log=False, low=8, step=8). See optuna/optuna#2941.
warnings.warn(message, FutureWarning)
C:\ProgramData\Anaconda3\lib\site-packages\optuna\distributions.py:766: FutureWarning: IntUniformDistribution(high=600, low=30, step=30) is deprecated and internally converted to IntDistribution(high=600, log=False, low=30, step=30). See optuna/optuna#2941.
warnings.warn(message, FutureWarning)
C:\ProgramData\Anaconda3\lib\site-packages\optuna\distributions.py:766: FutureWarning: IntUniformDistribution(high=50, low=5, step=5) is deprecated and internally converted to IntDistribution(high=50, log=False, low=5, step=5). See optuna/optuna#2941.
warnings.warn(message, FutureWarning)
C:\ProgramData\Anaconda3\lib\site-packages\optuna\distributions.py:766: FutureWarning: UniformDistribution(high=0.01, low=0.0001) is deprecated and internally converted to FloatDistribution(high=0.01, log=False, low=0.0001, step=None). See optuna/optuna#2941.
warnings.warn(message, FutureWarning)
== Status ==
Current time: 2022-12-27 14:21:07 (running for 00:00:05.18)
Memory usage on this node: 12.3/31.7 GiB
Using FIFO scheduling algorithm.
Resources requested: 0/20 CPUs, 0/1 GPUs, 0.0/12.44 GiB heap, 0.0/6.22 GiB objects (0.0/1.0 accelerator_type:G)
Result logdir: C:\Users\Steven\ray_results\run_trial_2022-12-27_14-21-02
Number of trials: 1/20 (1 PENDING)
+--------------------+----------+-------+--------------+----------------------+--------------+------------------------+------------+----------+
| Trial name | status | loc | batch_size | hidden_config | max_epochs | optimizer_params/lea | patience | use_bn |
| | | | |
| | rning_rate | | |
|--------------------+----------+-------+--------------+----------------------+--------------+------------------------+------------+----------|
| run_trial_41b3f5c7 | PENDING | | 104 | Choice_2: [64, _66c0 | 300 | 0.00969596 | 45 | False |
+--------------------+----------+-------+--------------+----------------------+--------------+------------------------+------------+----------+

2022-12-27 14:21:12,864 WARNING worker.py:1851 -- This worker was asked to execute a function that has not been registered ({type=PythonFunctionDescriptor, module_name=ray.util.placement_group, class_name=, function_name=_export_bundle_reservation_check_method_if_needed..bundle_reservation_check_func, function_hash=c0f5d1653dc84f95bbeed36cb66471cd}, node=127.0.0.1, worker_id=e226e520eafc567ab1ccf4e297e0559f5116718dbeda073aa17dee4a, pid=9048). You may have to restart Ray.
(pid=9048) 2022-12-27 14:21:12,863 ERROR function_manager.py:415 -- This worker was asked to execute a function that has not been registered ({type=PythonFunctionDescriptor, module_name=ray.util.placement_group, class_name=, function_name=_export_bundle_reservation_check_method_if_needed..bundle_reservation_check_func, function_hash=c0f5d1653dc84f95bbeed36cb66471cd}, node=127.0.0.1, worker_id=e226e520eafc567ab1ccf4e297e0559f5116718dbeda073aa17dee4a, pid=9048). You may have to restart Ray.
== Status ==
Current time: 2022-12-27 14:21:12 (running for 00:00:10.23)
Memory usage on this node: 12.3/31.7 GiB
Using FIFO scheduling algorithm.
Resources requested: 0/20 CPUs, 0/1 GPUs, 0.0/12.44 GiB heap, 0.0/6.22 GiB objects (0.0/1.0 accelerator_type:G)
Result logdir: C:\Users\Steven\ray_results\run_trial_2022-12-27_14-21-02
Number of trials: 1/20 (1 PENDING)
+--------------------+----------+-------+--------------+----------------------+--------------+------------------------+------------+----------+
| Trial name | status | loc | batch_size | hidden_config | max_epochs | optimizer_params/lea | patience | use_bn |
| | | | |
| | rning_rate | | |
|--------------------+----------+-------+--------------+----------------------+--------------+------------------------+------------+----------|
| run_trial_41b3f5c7 | PENDING | | 104 | Choice_2: [64, _66c0 | 300 | 0.00969596 | 45 | False |
+--------------------+----------+-------+--------------+----------------------+--------------+------------------------+------------+----------+

== Status ==
Current time: 2022-12-27 14:21:18 (running for 00:00:15.30)
Memory usage on this node: 12.3/31.7 GiB
Using FIFO scheduling algorithm.
Resources requested: 0/20 CPUs, 0/1 GPUs, 0.0/12.44 GiB heap, 0.0/6.22 GiB objects (0.0/1.0 accelerator_type:G)
Result logdir: C:\Users\Steven\ray_results\run_trial_2022-12-27_14-21-02
Number of trials: 1/20 (1 PENDING)
+--------------------+----------+-------+--------------+----------------------+--------------+------------------------+------------+----------+
| Trial name | status | loc | batch_size | hidden_config | max_epochs | optimizer_params/lea | patience | use_bn |
| | | | |
| | rning_rate | | |
|--------------------+----------+-------+--------------+----------------------+--------------+------------------------+------------+----------|
| run_trial_41b3f5c7 | PENDING | | 104 | Choice_2: [64, _66c0 | 300 | 0.00969596 | 45 | False |
+--------------------+----------+-------+--------------+----------------------+--------------+------------------------+------------+----------+

运行环境

windows 11 vscode powershell
paddlepaddle-gpu 2.4.1
paddlets 1.0.2

您好，请教一个问题，现在有一系列无标签的时序数据，也不清楚这些数据有几个类别，请问PaddleTS是否有模型能够对这样的数据进行聚类？是否有相关的示例代码？

这个例子中，假如 id的顺序 是0，1交叉的，就报错了。不支持这种吗？

sample = pd.DataFrame(np.random.randn(200, 3), columns=['a', 'c', 'd'])
sample['id'] = pd.Series([0]*80 + [1]*120, name='id')

#Load TSDatasets by group_id
from paddlets import TSDataset
tsdatasets = TSDataset.load_from_dataframe(
df=sample,
group_id='id',
target_cols='a',
observed_cov_cols=['c', 'd'],
#static_cov_cols='id'
)

报错具体信息：ValueError: known_cov length is too short to build known_cov chunk feature.
It needs at least 1 extra Timestamps after known_timeseries.time_index[6:]
数据集构建：
custom_dataset = TSDataset.load_from_dataframe(
result, #Also can be path to the CSV file
time_col='time',
target_cols=['A' ,'B' ,'C'],
freq='1d',
observed_cov_cols='D',
fill_missing_dates=True,
fillna_method='zero' #max, min, avg, median, pre, back, zero
)
custom_dataset.plot()

我是一个时序数据方面的新手，通过官方文档load自己的数据(数据时间间隔是不固定的)，输出的结果如下图所示。经过阅读源码发现是freq参数没有理解到位。我觉得pandas中的asfreq()方法其他参数也应该提供用户使用(例如：method和fill_value)这样可以让我这样的新手避免出现数据全是NAN的情况。

1.千呼万唤始出来,终于有了飞桨框架的时间序列预测了.
2.我进入时间序列预测领域,都是由于获得AAAI2021最佳论文奖的那篇论文引起我的兴趣的.AAAI 2021最佳论文：比Transformer更有效的长时间序列预测.看到这个标题,竟然比Transformer更牛逼,万分吃惊.请问PaddleTS大佬,几时复现这篇论文呢?或者是复现2022年比较牛逼的时间序列预测的论文.如果我记得不错的话,AAAI2021最佳论文,是有pytorch的复现版本的,就在github.其实多数论文都有pytorch的复现版本,毕竟科研者写论文,超过80%以上的论文就是使用pytorch框架.使用飞桨框架复现,可以参考参考.
3.看到大佬们对于未来的工作的概括,小弟最感兴趣的就是概率模型,希望早日复现该领域最牛逼的论文.我听说有人使用概率模型炒股,发了大财,不知道真假,哈哈哈哈.
4.图像分类的数据集,有一个很大的imagenet数据集,请问时间序列预测,有什么很大的数据集吗?目前PaddleTS提供了很多模型,请问你们认为哪个模型效果最好呢?哪个模型适用于短时间序列预测,哪个模型适用于长时间序列预测?哪个模型适用于小数据集,哪个模型适用于大数据集?
5.我是看了飞桨视觉的负责人的一篇微信朋友圈,才知道PaddleTS这个仓库.朋友圈中介绍,时间序列预测可以用于股价预测,请问你们认同时间序列预测可以用于股价预测这种观点吗?祝你们炒股赚大钱,发大财.

1.支持目标变量分类,先实现所有模型和算法,单目标变量分类功能,再实现多目标变量都是整型数的功能,再实现多目标变量有的是整形术,有的是浮点数的功能.
2.支持单机多卡GPU训练,能支持分布式训练就更好了,但是单机多卡训练是最重要的,很多人没钱买机器,租用八卡3090ti进行训练,真心爽歪歪.对于paddlets来说,如果已经选定模型,设置好超参数,或许不用太多GPU资源,但是如果刚开始一个项目,所有模型都要测试一遍,甚至超参数都要多测试,单机多卡GPU训练是很重要的,一个模型训练1天,测试一遍所有模型+超参数那就是一个月的事情了,慢啊.
3.支持C++推理,我指的是,使用VS2022,C++20标准写代码,能准确运行程序.当然了,支持CUDA,OPENCL也是需求之一,这样做可以大幅度提高训练和推理的速度吧.AMD的显卡没什么需求,算了吧,真没想到,谁会买AMD显卡.
4.在ai studio加点项目或案例
5.能不能弄个大数据集.
6.我之前花时间,看了paddlets所有文档,没看代码.给我的感觉就是,预测目标变量的未来值,基本依靠目标变量的过去值,协变量只是起到辅助作用,如果目标变量的未来值,就是靠协变量预测,跟目标变量的过去值没什么关系,还能用paddlets吗?比如y=2*x,y的未来值,是由协变量x决定的,跟y的过去值毫无关系,这时候,能用paddlets吗?
7.我没看过paddlets的模型代码,但我估计这个领域的模型都不大,换句话说,不太清楚paddleslim是否有用.但以我的经验,PaddleSlim用于paddlets,还是能降低使用者推理成本的,比如原来推理需要3090ti,就可以换成2070.拿图像分类来说,有些图像分类不难,只使用很小的神经网络就可以达到99准确率,用了paddleclas的最小模型进行分类的训练,然后用PaddleSlim剪枝量化等,生产出来的模型,相同机器下,推理速度大幅提升,或者是,推理速度不变,GPU,CPU成本大幅下降.总结,paddlets支持PaddleSlim还是很有用的,哈哈哈哈.

PaddlePaddle官方已经支持 Mac M1芯片的运行。但是，PaddleTS却不能支持。
我看到PaddleTS中numpy依赖的版本是1.19.5，但是1.19.5并没有macOS的arm64的版本。

在#issue280中，通过相关部分修改能够成功运行，但是我发现运行结果绘图时，正式值和原来的真实值出现了偏差。
这是默认版本的deeper绘制的效果图

然后我使用了自动调参,获取到了一个最优参数，然后绘制的结果图如下：

这里可以看到truth值，也就是真实值都发生了变化，和原来的真实值不符合，请问这是什么情况呢？

相关环境如下：

Name Version Build Channel

aiosignal 1.2.0 pypi_0 pypi
alembic 1.8.1 pypi_0 pypi
amqp 5.1.1 pypi_0 pypi
anyio 3.6.2 pypi_0 pypi
argon2-cffi 21.3.0 pypi_0 pypi
argon2-cffi-bindings 21.2.0 pypi_0 pypi
astor 0.8.1 pypi_0 pypi
asttokens 2.1.0 pypi_0 pypi
attrs 22.1.0 pypi_0 pypi
autopage 0.5.1 pypi_0 pypi
backcall 0.2.0 pypi_0 pypi
beautifulsoup4 4.11.1 pypi_0 pypi
billiard 3.6.4.0 pypi_0 pypi
bleach 5.0.1 pypi_0 pypi
ca-certificates 2022.07.19 haa95532_0 https://mirrors.aliyun.com/anaconda/pkgs/main
celery 5.2.7 pypi_0 pypi
certifi 2022.9.14 py38haa95532_0 https://mirrors.aliyun.com/anaconda/pkgs/main
cffi 1.15.1 pypi_0 pypi
charset-normalizer 2.1.1 pypi_0 pypi
chinese-calendar 1.8.0 pypi_0 pypi
click 8.0.4 pypi_0 pypi
click-didyoumean 0.3.0 pypi_0 pypi
click-plugins 1.1.1 pypi_0 pypi
click-repl 0.2.0 pypi_0 pypi
cliff 4.0.0 pypi_0 pypi
cloudpickle 2.2.0 pypi_0 pypi
cmaes 0.8.2 pypi_0 pypi
cmd2 2.4.2 pypi_0 pypi
colorama 0.4.5 pypi_0 pypi
colorlog 6.7.0 pypi_0 pypi
configspace 0.6.0 pypi_0 pypi
contourpy 1.0.5 pypi_0 pypi
cycler 0.11.0 pypi_0 pypi
cython 0.29.32 pypi_0 pypi
debugpy 1.6.3 pypi_0 pypi
decorator 5.1.1 pypi_0 pypi
defusedxml 0.7.1 pypi_0 pypi
distlib 0.3.6 pypi_0 pypi
entrypoints 0.4 pypi_0 pypi
executing 1.2.0 pypi_0 pypi
fastjsonschema 2.16.2 pypi_0 pypi
filelock 3.8.0 pypi_0 pypi
flaml 1.0.12 pypi_0 pypi
flask 2.2.2 pypi_0 pypi
flask-cors 3.0.10 pypi_0 pypi
fonttools 4.37.3 pypi_0 pypi
frozenlist 1.3.1 pypi_0 pypi
greenlet 1.1.3 pypi_0 pypi
grpcio 1.43.0 pypi_0 pypi
hpbandster 0.7.4 pypi_0 pypi
idna 3.4 pypi_0 pypi
importlib-metadata 4.12.0 pypi_0 pypi
importlib-resources 5.9.0 pypi_0 pypi
ipykernel 6.17.0 pypi_0 pypi
ipython 8.6.0 pypi_0 pypi
ipython-genutils 0.2.0 pypi_0 pypi
ipywidgets 8.0.2 pypi_0 pypi
itsdangerous 2.1.2 pypi_0 pypi
jedi 0.18.1 pypi_0 pypi
jinja2 3.1.2 pypi_0 pypi
joblib 1.2.0 pypi_0 pypi
jsonschema 4.16.0 pypi_0 pypi
jupyter 1.0.0 pypi_0 pypi
jupyter-client 7.4.4 pypi_0 pypi
jupyter-console 6.4.4 pypi_0 pypi
jupyter-core 4.11.2 pypi_0 pypi
jupyter-server 1.21.0 pypi_0 pypi
jupyterlab-pygments 0.2.2 pypi_0 pypi
jupyterlab-widgets 3.0.3 pypi_0 pypi
kiwisolver 1.4.4 pypi_0 pypi
kombu 5.2.4 pypi_0 pypi
lightgbm 3.3.2 pypi_0 pypi
llvmlite 0.39.1 pypi_0 pypi
lxml 4.9.1 pypi_0 pypi
mako 1.2.2 pypi_0 pypi
markupsafe 2.1.1 pypi_0 pypi
matplotlib 3.6.0 pypi_0 pypi
matplotlib-inline 0.1.6 pypi_0 pypi
mistune 2.0.4 pypi_0 pypi
more-itertools 8.14.0 pypi_0 pypi
mpmath 1.2.1 pypi_0 pypi
msgpack 1.0.4 pypi_0 pypi
nbclassic 0.4.8 pypi_0 pypi
nbclient 0.7.0 pypi_0 pypi
nbconvert 7.2.3 pypi_0 pypi
nbformat 5.7.0 pypi_0 pypi
nest-asyncio 1.5.6 pypi_0 pypi
netifaces 0.11.0 pypi_0 pypi
notebook 6.5.2 pypi_0 pypi
notebook-shim 0.2.2 pypi_0 pypi
numba 0.56.4 pypi_0 pypi
numpy 1.19.5 pypi_0 pypi
openssl 1.1.1q h2bbff1b_0 https://mirrors.aliyun.com/anaconda/pkgs/main
opt-einsum 3.3.0 pypi_0 pypi
optuna 3.0.3 pypi_0 pypi
packaging 21.3 pypi_0 pypi
paddle-bfloat 0.1.7 pypi_0 pypi
paddlepaddle-gpu 2.3.2 pypi_0 pypi
paddlets 1.1.0 pypi_0 pypi
pandas 1.3.5 pypi_0 pypi
pandocfilters 1.5.0 pypi_0 pypi
parso 0.8.3 pypi_0 pypi
patsy 0.5.2 pypi_0 pypi
pbr 5.10.0 pypi_0 pypi
pickleshare 0.7.5 pypi_0 pypi
pillow 9.2.0 pypi_0 pypi
pip 22.1.2 py38haa95532_0 https://mirrors.aliyun.com/anaconda/pkgs/main
pkgutil-resolve-name 1.3.10 pypi_0 pypi
platformdirs 2.5.2 pypi_0 pypi
prettytable 3.4.1 pypi_0 pypi
prometheus-client 0.15.0 pypi_0 pypi
prompt-toolkit 3.0.32 pypi_0 pypi
protobuf 3.20.0 pypi_0 pypi
psutil 5.9.3 pypi_0 pypi
pure-eval 0.2.2 pypi_0 pypi
pycparser 2.21 pypi_0 pypi
pygments 2.13.0 pypi_0 pypi
pyod 1.0.6 pypi_0 pypi
pyparsing 3.0.9 pypi_0 pypi
pyperclip 1.8.2 pypi_0 pypi
pyreadline3 3.4.1 pypi_0 pypi
pyro4 4.82 pypi_0 pypi
pyrsistent 0.18.1 pypi_0 pypi
python 3.8.13 h6244533_0 https://mirrors.aliyun.com/anaconda/pkgs/main
python-dateutil 2.8.2 pypi_0 pypi
python-docx 0.8.11 pypi_0 pypi
pytz 2022.2.1 pypi_0 pypi
pywavelets 1.3.0 pypi_0 pypi
pywin32 304 pypi_0 pypi
pywinpty 2.0.9 pypi_0 pypi
pyyaml 6.0 pypi_0 pypi
pyzmq 24.0.1 pypi_0 pypi
qtconsole 5.4.0 pypi_0 pypi
qtpy 2.2.1 pypi_0 pypi
ray 2.1.0 pypi_0 pypi
requests 2.28.1 pypi_0 pypi
scikit-learn 1.1.2 pypi_0 pypi
scipy 1.7.3 pypi_0 pypi
seaborn 0.12.1 pypi_0 pypi
send2trash 1.8.0 pypi_0 pypi
serpent 1.41 pypi_0 pypi
setuptools 63.4.1 py38haa95532_0 https://mirrors.aliyun.com/anaconda/pkgs/main
shap 0.41.0 pypi_0 pypi
six 1.16.0 pypi_0 pypi
slicer 0.0.7 pypi_0 pypi
sniffio 1.3.0 pypi_0 pypi
soupsieve 2.3.2.post1 pypi_0 pypi
sqlalchemy 1.4.41 pypi_0 pypi
sqlite 3.39.2 h2bbff1b_0 https://mirrors.aliyun.com/anaconda/pkgs/main
stack-data 0.6.0 pypi_0 pypi
statsmodels 0.12.2 pypi_0 pypi
stevedore 4.0.0 pypi_0 pypi
sympy 1.11.1 pypi_0 pypi
tabulate 0.8.10 pypi_0 pypi
tensorboardx 2.5.1 pypi_0 pypi
terminado 0.17.0 pypi_0 pypi
threadpoolctl 3.1.0 pypi_0 pypi
tinycss2 1.2.1 pypi_0 pypi
tornado 6.2 pypi_0 pypi
tqdm 4.64.1 pypi_0 pypi
traitlets 5.5.0 pypi_0 pypi
typing-extensions 4.3.0 pypi_0 pypi
urllib3 1.26.12 pypi_0 pypi
vc 14.2 h21ff451_1 https://mirrors.aliyun.com/anaconda/pkgs/main
vine 5.0.0 pypi_0 pypi
virtualenv 20.16.5 pypi_0 pypi
vs2015_runtime 14.27.29016 h5e58377_2 https://mirrors.aliyun.com/anaconda/pkgs/main
wcwidth 0.2.5 pypi_0 pypi
webencodings 0.5.1 pypi_0 pypi
websocket-client 1.4.2 pypi_0 pypi
werkzeug 2.2.2 pypi_0 pypi
wheel 0.37.1 pyhd3eb1b0_0 https://mirrors.aliyun.com/anaconda/pkgs/main
widgetsnbextension 4.0.3 pypi_0 pypi
wincertstore 0.2 py38haa95532_2 https://mirrors.aliyun.com/anaconda/pkgs/main
xgboost 1.6.2 pypi_0 pypi
zipp 3.8.1 pypi_0 pypi

报错部分代码：

import paddle
import numpy as np

固定随机随机种子，保证训练结果可复现

seed = 2022
paddle.seed(seed)
np.random.seed(seed)

#建模与训练
from paddlets.models.anomaly import AutoEncoder
model = AutoEncoder(
in_chunk_len=2, # 样本数据窗口大小
max_epochs=100 # 最大epoch设为100
)
model.fit(train_data_scaled)

#报错内容如下

ModuleNotFoundError Traceback (most recent call last)
Cell In [6], line 10
7 np.random.seed(seed)
9 #建模与训练
---> 10 from paddlets.models.anomaly import AutoEncoder
11 model = AutoEncoder(
12 in_chunk_len=2, # 样本数据窗口大小
13 max_epochs=100 # 最大epoch设为100
14 )
15 model.fit(train_data_scaled)

File E:\Paddle-release-2.2\PaddleTS\paddlets\models\anomaly_init_.py:8
1 # !/usr/bin/env python3
2 # -- coding:utf-8 --
4 """
5 paddlets anomaly.
6 """
----> 8 from paddlets.models.anomaly.dl.autoencoder import AutoEncoder
9 from paddlets.models.anomaly.dl.anomaly_transformer import AnomalyTransformer
10 from paddlets.models.anomaly.dl.vae import VAE

File E:\Paddle-release-2.2\PaddleTS\paddlets\models\anomaly\dl\autoencoder.py:15
13 from paddlets.models.anomaly.dl._ed.ed import MLP, CNN
14 from paddlets.models.common.callbacks import Callback
---> 15 from paddlets.models.anomaly.dl import utils as U
16 from paddlets.datasets import TSDataset
17 from paddlets.logger import raise_if, raise_if_not

File E:\Paddle-release-2.2\PaddleTS\paddlets\models\anomaly\dl\utils.py:11
9 import paddle
10 import paddle.nn.functional as F
---> 11 import more_itertools as mit
13 from paddlets.logger import raise_if_not, raise_if, raise_log, Logger
14 from paddlets.datasets import TSDataset

ModuleNotFoundError: No module named 'more_itertools'