xy-chen16 / epianno Goto Github PK

Dear authors,

I find your paper very interesting to follow. I appreciate that you release the code.
However, when I use it with my own data, it throws the following exceptions:

 File "run_crossdataset_projection.py", line 72, in <module>
    pmu,psigma,pz,pw,pnoise,px = EpiAnno.EpiAnno(10,data_train.shape[1],n_classes,sample_shape)
File "xxx/benchmark_EpiAnno/code/EpiAnno.py", line 29, in EpiAnno
    x = ed.Normal(loc = output, scale = noise, name = 'x')
...
lib/python3.7/site-packages/tensorflow_core/python/framework/tensor_util.py", line 520, in make_tensor_proto
    "Cannot create a tensor proto whose content is larger than 2GB.")
ValueError: Cannot create a tensor proto whose content is larger than 2GB.

For x = ed.Normal(loc = output, scale = noise, name = 'x'), input argument information is: output shape=(8150, 72822), noise shape=(1,)

I highly suspect it might be caused by my large input data,

total 21G
-rw-r--r-- 1 xxxx nts 126K Mar 21 22:37 ref-immunePBMC_PBMC_Rep1-cell_type.hkl
-rw-r--r-- 1 xxxx nts  13G Mar 21 22:44 ref-immunePBMC_PBMC_Rep1-data_count.hkl
-rw-r--r-- 1 xxxx nts 8.7G Mar 21 22:45 tgt-immunePBMC_ref_PBMC_Rep1_tgt_PBMC_Rep2-data_count.hkl

I was wondering by any chance you may know the solution or workaround for it?

Thanks!

Hello
When I run EpiAnno using the code (python run_self_projection.py -d ../data/InSilico_count.hkl -c ../data/InSilico_cell_type.hkl -o outdir),(without -g gpu), only get some .npy files in outdir. And the last STDout is

VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray

Do you know how to fix it.

I could not see the quick-start notebook.
Could you fix it.

Sincerely
Xiaolin