PERF-#7230: Don't preserve bad partition for merge

[anmyachev]

What do these changes do?

merge preserves the row splitting, but sometimes it's better to repartition. One extremely bad case is when there is a sequence of heavyweight operations where inefficient partitioning persists from the first to the last. For example, sequence of merge operations, where the left operand of the very first operation in the chain has only one partition. For example:

import modin.pandas as pd
import numpy as np
from time import time
from modin.utils import execute
import modin.config as cfg

cfg.NPartitions.put(8)


pd.DataFrame(np.arange(cfg.NPartitions.get() * cfg.MinPartitionSize.get())).to_numpy()

for _ in range(3):
    np.random.seed(42)
    # small
    df1 = pd.DataFrame(np.random.randint(100, size=(32, 16)))
    assert df1._query_compiler._modin_frame._partitions.shape == (1, 1)

    # big
    df2 = pd.DataFrame(np.random.randint(100, size=(16_000, 16)))
    df3 = pd.DataFrame(np.random.randint(100, size=(200_000, 16)))

    start = time()

    res = df1.merge(df2, on=2)
    res = res.merge(df3, left_on='3_x', right_on=3)
    print(res.shape)
    execute(res)

    print(f"time: {time()-start}")

Results: 5.07 sec (on main) vs 2.56 sec (in the PR)

• first commit message and PR title follow format outlined here

  NOTE: If you edit the PR title to match this format, you need to add another commit (even if it's empty) or amend your last commit for the CI job that checks the PR title to pick up the new PR title.

• passes flake8 modin/ asv_bench/benchmarks scripts/doc_checker.py
• passes black --check modin/ asv_bench/benchmarks scripts/doc_checker.py
• signed commit with git commit -s
• Resolves Don't preserve bad partition for merge #7230
• tests added and passing
• module layout described at docs/development/architecture.rst is up-to-date

[YarShev]

What do we test here?

[anmyachev]

That in case of bad partitioning repartition was called (when performing merge operation).

[YarShev]

Does this change slow down some of our benchmarks?