read_excel fails if usecols and index_cols parameters are provided

[amyskov]

System information

• OS Platform and Distribution (e.g., Linux Ubuntu 16.04): Ubuntu 20.04
• Modin version: 54898ef
• Python version: 3.8.10
• Code we can use to reproduce:

import pandas
import numpy as np
import os

import modin.pandas as pd
from modin.pandas.test.utils import df_equals
from modin.config import NPartitions
NPartitions.put(4)

filename = "test_excel.xlsx"
row_size = 256
kwargs = {
    "usecols": [0],
    "index_col": 0,
}


try:
    df = pandas.DataFrame(
        {"col1": np.arange(row_size), "col2": np.arange(row_size)}
    )
    df.to_excel(filename)

    df_pd = pd.read_excel(filename, **kwargs)
    df_pandas = pandas.read_excel(filename, **kwargs)

    df_equals(df_pandas, df_pd)

finally:
    os.remove(filename)

Describe the problem

Source code / logs

Traceback (most recent call last):
  File "test.py", line 824, in <module>
    df_pd = pd.read_excel(filename, **kwargs)
  File "/modin/modin/pandas/io.py", line 347, in read_excel
    intermediate = FactoryDispatcher.read_excel(**kwargs)
  File "/modin/modin/data_management/factories/dispatcher.py", line 202, in read_excel
    return cls.__factory._read_excel(**kwargs)
  File "/modin/modin/data_management/factories/factories.py", line 256, in _read_excel
    return cls.io_cls.read_excel(**kwargs)
  File "/modin/modin/engines/base/io/file_dispatcher.py", line 67, in read
    query_compiler = cls._read(*args, **kwargs)
  File "/modin/modin/engines/base/io/text/excel_dispatcher.py", line 246, in _read
    dtypes = pandas.Series(dtypes, index=column_names)
  File "/miniconda3/envs/modin/lib/python3.8/site-packages/pandas/core/series.py", line 355, in __init__
    if is_empty_data(data) and dtype is None:
  File "/miniconda3/envs/modin/lib/python3.8/site-packages/pandas/core/construction.py", line 794, in is_empty_data
    is_simple_empty = is_list_like_without_dtype and not data
  File "/miniconda3/envs/modin/lib/python3.8/site-packages/pandas/core/generic.py", line 1534, in __nonzero__
    raise ValueError(
ValueError: The truth value of a DataFrame is ambiguous. Use a.empty, a.bool(), a.item(), a.any() or a.all().

[RehanSD]

Confirmed that this bug is reproducible on master, although the error is different:

AssertionError                            Traceback (most recent call last)
Input In [1], in <cell line: 18>()
     24     df_pd = pd.read_excel(filename, **kwargs)
     25     df_pandas = pandas.read_excel(filename, **kwargs)
---> 27     df_equals(df_pandas, df_pd)
     29 finally:
     30     os.remove(filename)

File ~/Documents/modin/modin/pandas/test/utils.py:583, in df_equals(df1, df2)
    580     df2 = to_pandas(df2)
    582 if isinstance(df1, pandas.DataFrame) and isinstance(df2, pandas.DataFrame):
--> 583     assert_empty_frame_equal(df1, df2)
    585 if isinstance(df1, pandas.DataFrame) and isinstance(df2, pandas.DataFrame):
    586     assert_frame_equal(
    587         df1,
    588         df2,
   (...)
    593         check_categorical=False,
    594     )

File ~/Documents/modin/modin/pandas/test/utils.py:528, in assert_empty_frame_equal(df1, df2)
    513 """
    514 Test if df1 and df2 are empty.
    515
   (...)
    524     If check fails.
    525 """
    527 if (df1.empty and not df2.empty) or (df2.empty and not df1.empty):
--> 528     assert False, "One of the passed frames is empty, when other isn't"
    529 elif df1.empty and df2.empty and type(df1) != type(df2):
    530     assert False, f"Empty frames have different types: {type(df1)} != {type(df2)}"

AssertionError: One of the passed frames is empty, when other isn't