Bug Report

These tests were run on s390x. s390x is big-endian architecture.

Failure log for helper_test.py

________________________________________________ TestHelperTensorFunctions.test_make_tensor ________________________________________________

self = <helper_test.TestHelperTensorFunctions testMethod=test_make_tensor>

    def test_make_tensor(self):  # type: () -> None
    

When merging a dask dataframe, the resulting index is duplicated - seems to be because of the number of partitions. See example below:

import pandas as pd
import dask.dataframe as dd

a = dd.from_pandas(pd.DataFrame({'a': [1,2,3,4]}), npartitions=2)
b = pd.DataFrame({'a': [1,2,3,4], 'b': [2,3,4,5]})

a.merge(b, on='a').compute()

Returns

|

For example, if there is a relationship transaction.session_id -> sessions.id and we are calculating a feature transactions: sessions.SUM(transactions.value) any rows for which there is no corresponding session should be given the default value of 0 instead of NaN.

Of course this should not normally occur, but when it does it seems more reasonable to use the default_value.

`DirectF

with the Power Transformer.

I see the code

device = ‘cuda’ if torch.cuda.is_available() else ‘cpu’

repeated often in user code. Maybe we should introduce device='auto' exactly for this case?

Yes

How do i resume training for text classification?

Describe the solution you'd like
We already have AutoARIMA, but it would be nice to also interface ARIMA, either from statsmodels or pmdarima.

I think it could be useful, when one wants to plot only e.g. class 1, to have an option to produce consistent plots for both plot_cumulative_gain and plot_roc

At the moment, instead, only plot_roc supports such option.

Thanks a lot

Support Series.median()