When I ran tests.py, I've this strange error. It occurred from the first line of the CoreTest.test_type test case.

It seems that the ouput dict contains some integers which are integers of type np.int64. This numpy int type is not JSON serializable. Call serialize on your DataFrame from the tests.py file with the output_type="json"`parameter and do:

import json
# 'df' from tests.py
json_data = serialize(df, render_to="chart", output_type="json")
# json.dumps(json_data) # raises the TypeError
type(json_data["series"][0]["data"][0][0])
# => numpy.int64

Unfortunately, when I read the code, the serialize_series function specifically, I cannot see where you can have such type.

I wonder whether I can set the figure size? Sometimes the figure is two narrow in the jupyter.

Hello,

maybe you should put in doc a sample DataFrame

with

dat = """ts;A;B;C
2015-01-01 00:00:00;27451873;29956800;113
2015-01-01 01:00:00;20259882;17906600;76
2015-01-01 02:00:00;11592256;12311600;48
2015-01-01 03:00:00;11795562;11750100;50
2015-01-01 04:00:00;9396718;10203900;43
2015-01-01 05:00:00;14902826;14341100;53"""
from pandas.compat import StringIO
df = pd.read_csv(StringIO(dat), sep=';', index_col='ts', parse_dates='ts')

or with a random walk

N = 600
index = pd.date_range('2015-01-01', periods=N, freq='1D')
df = pd.DataFrame(np.random.randn(N, 3), index=index, columns=list('ABC'))
df = df.cumsum() + 800

it will be easier to give pandas-highcharts a try.

Here is a link to show it
http://nbviewer.ipython.org/github/scls19fr/pandas_timeseries_notebooks_samples/blob/master/pandas-highcharts.ipynb

Kind regards

I see that the last commit was in 2017, is it still active?

Thanks,
Eli

I am having issues adding PlotLines onto the charts after rendering into my template.

This code adds a plotline:

chart.xAxis[0].addPlotLine({
                   value: 1000,
                   color: 'red',
                   width: 2,
                   id: 'plot-line-1',
                   zIndex: 3,
 });

The plot lines are being added to the html, but with some funky characteristics:

<g class="highcharts-plot-lines-0">
    <path fill="none" stroke="red" stroke-width="200" visibility="hidden"></path>
</g>

Namely:

1. Visibility is set to "hidden"
2. There is no 'd' attribute in the path

Is this likely to be a Highcharts failure?

I was thinking about refactoring your code a little bit by making a Highchart object as in class Highchart(object): and would make the serialize function use this object and behave the exact same way it does at the moment.

I want to do that to be able to generate the object (without using the serialize function), alter the data structure at will, and then render it my way.

Would you review / accept such a PR?

Hi ,

I downloaded this package and installed it for Python 3.4. it did not work.

I changed core.py module to

• Work with Python 3
• Made APIs more generic cab called by pandas or non-panda by passing X_LIST and Y_LIST to be plotted as two separate list.
• Added support for 'pie' plot.

Do you want me to submit code to you so that yo can make new version.
If not is it okay to create branch 'yieldfrom-pandas-highchart' and add changes there ?

Regards,
Shivaram

Used in Jupyterlab, pandas_highcharts figures do not render. Repro in Jupyterlab:

[1]  import pandas as pd;
from pandas_highcharts.display import display_charts

df = pd.read_csv(...)
display_charts(df, title='Title')

Output is simply an empty cell.

I expect this is related to Jupyterlab's restriction of Javascript execution and requirement of a Jupyterlab extension: jupyterlab/jupyterlab#3748

Like column charts are not. Package is easy to use given the dataframe support but would save time to know what chart types are actually supported

ref: http://www.highcharts.com/stock/demo & http://api.highcharts.com/highstock

1. Thank you very much for your work. Combination of pandas with highcharts makes the graphic creation on the webpage a very entertaining exercise.

Nevertheless under OSX I have notices a problem when importing the display_charts.
Everything appears to work as expected when using the IPython (v4.1.2). When trying to use the examples from pure python (v2.7.10) I see the following error:

# > python  app2.py

Traceback (most recent call last):
  File "app2.py", line 11, in <module>
    from pandas_highcharts.display import display_charts
  File "build/bdist.macosx-10.11-x86_64/egg/pandas_highcharts/display.py", line 32, in <module>
AttributeError: 'NoneType' object has no attribute 'config'

The app2.py code presents itself as:

from flask import Flask, render_template
import pandas as pd
from pandas_highcharts.core import serialize
import ConfigParser
import os

from pandas_highcharts.display import display_charts
[..]

Am I missing something ?

Thank you in advance for your comments / support.

Cheers,
ManieqZ

Hello,

How to use the Column with drill down from Highcharts with a multilevel indexed data frame ?

http://www.highcharts.com/demo/column-drilldown

Hey there, I'm trying to download the example notebooks with the plots as HTML and the charts won't show - there are a bunch of ERR_FILE_NOT_FOUND errors in the JS console.

Am I missing an intermediary step to be able to do this?

Cheers,
Sam

Python 2.7.7 32bits on Windows 7 with the Anaconda distribution (I don't like this OS for dev but, it's for work you know)

Raised OverflowError: mktime argument out of range

• http://stackoverflow.com/questions/2518706/python-mktime-overflow-error
• https://docs.python.org/2/library/time.html#time.mktime

If the input value cannot be represented as a valid time, either OverflowError or ValueError will be
raised (which depends on whether the invalid value is caught by Python or the underlying C libraries

Hello,

I installed pandas-highcharts using

$ pip install git+https://github.com/gtnx/pandas-highcharts.git

but

import pandas_highcharts
pandas_highcharts.serialize

raises

AttributeError: 'module' object has no attribute 'serialize'

Any idea ?

https://pypi.python.org/pypi/tox

Hi,
Thank you for providing this useful tool.
I am using it for plotting bar with line and have some problem with it. But seems it will show in two separate graph rather than one graph with two yaxis. My code is like this
display_charts(df['A'], kind="bar", title="Germany inflation rate")
display_charts(df['B'], kind="line", title="Germany inflation rate",secondary_y = [True])
So could you help solve this problem?

Thanks!

Is this on the roadmap or out of scope?

In general it would be good to have a lower level control of the chart

I tried to modify your code but couldn't make it work.. Trying to add this chart: http://jsfiddle.net/gh/get/jquery/3.1.1/highslide-software/highcharts.com/tree/master/samples/stock/demo/ohlc/

Any idea on what to modify? Thanks! Great library btw!

Hello,

I noticed some minor fixes to sample notebook for Python 3 (Pandas Highcharts Example - example.ipynb)

from StringIO import StringIO

should be

from pandas.compat import StringIO

and

from urllib2 import urlopen

should be

from pandas.io.common import urlopen

in display.py

from core import serialize

should be changed to

from .core import serialize

xrange should also be changed to range with

from pandas.compat import range

Kind regards

I would like use something title.
Like this dict style.

{
    "text": "Chart",
    "x": -20 #Center
}

I replicated the example notebook step by step, and while the chart renders perfectly - the top-right icon to export the chart in various file options (which is a default for highcharts) doesn't appear. All the notebook examples show that option to be present. Have there been some changes recently?

• the serialize function raised a JSONEncoder error for your first chart example (data from Quandl)
  (I use pandas 0.15.2 but I don't know if it's related)
• If you convert the string date representation from Quandl data with pd.read_csv([...], parse_dates=[0]) (suppose "Date" is the first column), your serializer works very well and turns dates into a JSON format that Highchart can understand (you don't have to convert string into a int I guess)
• The first display(HTML("""[highchart <script>]""")) cell have a%` but nothing to replace in the string.
• I wonder why you call locals() instead of a dict with a simple key,value with your chart.
• Good job!!

Hi,

Is is possible to have git tags in the repo which follow the version of the code on PyPi?

Thanks,
Damien

compare and compareBase not supported yet

https://api.highcharts.com/highstock/plotOptions.series.compare

I think you need at least from core import serialize in init

Else I cannot use pandas_highcharts.serialize

like this http://www.highcharts.com/stock/demo/dynamic-update

Hi,

This is great work. Are you planning to add any additional chart types and some of the other HC features any time soon? I'm particularly interested in rendering some big data charts that will need to look a bit like this.

http://jsfiddle.net/ttqemvzs/

Is there a template that can be edited fairly easily to add custom charts ?

Hi,

Pretty neat your Python package. I wanted to start a similar tool from scratch (pandas conversion to JSON data as input params for Highcharts) and I find out yours. Very nice !

Cheers,
Damien

I am told to pass 'ts' as booleans, Lists and dictionaries. How do I resolve this error.