sktime: [BUG] dependency bug in tsfresh causing tests to fail

Suggested fix:

• set an upper bound on the last statsmodels version that still worked in setup.py, see #1478
• same for the binder/build requirements
• wait for tsfresh to resolve the issue
• remove the version bound on statsmodels and test whether things run again

---

The CI tests are breaking, due to a tsfresh import now fails, which is required for TSfreshClassifier.

Error as below.

Looks like this is coming from an interface change in statsmodels, as a tsfresh dependency? Where tsfresh is calling stattools.acf which no longer has an unbiased argument in 0.13.0, but used to have?

Not sure how to best fix this, it seems it needs to be fixed in tsfresh? Can someone check whether the diagnosis is correct?

070     >>> from sktime.classification.feature_based import TSFreshClassifier
071     >>> from sktime.datasets import load_italy_power_demand
072     >>> X_train, y_train = load_italy_power_demand(split="train", return_X_y=True)
073     >>> X_test, y_test = load_italy_power_demand(split="test", return_X_y=True)
074     >>> clf = TSFreshClassifier()
075     >>> clf.fit(X_train, y_train)
UNEXPECTED EXCEPTION: TypeError("acf() got an unexpected keyword argument 'unbiased'")
multiprocessing.pool.RemoteTraceback: 
"""
Traceback (most recent call last):
  File "/opt/_internal/cpython-3.7.12/lib/python3.7/multiprocessing/pool.py", line 121, in worker
    result = (True, func(*args, **kwds))
  File "/opt/_internal/cpython-3.7.12/lib/python3.7/site-packages/tsfresh/utilities/distribution.py", line 43, in _function_with_partly_reduce
    results = list(itertools.chain.from_iterable(results))
  File "/opt/_internal/cpython-3.7.12/lib/python3.7/site-packages/tsfresh/utilities/distribution.py", line 42, in <genexpr>
    results = (map_function(chunk, **kwargs) for chunk in chunk_list)
  File "/opt/_internal/cpython-3.7.12/lib/python3.7/site-packages/tsfresh/feature_extraction/extraction.py", line 337, in _do_extraction_on_chunk
    return list(_f())
  File "/opt/_internal/cpython-3.7.12/lib/python3.7/site-packages/tsfresh/feature_extraction/extraction.py", line 317, in _f
    result = func(x, param=parameter_list)
  File "/opt/_internal/cpython-3.7.12/lib/python3.7/site-packages/tsfresh/feature_extraction/feature_calculators.py", line 404, in agg_autocorrelation
    a = acf(x, unbiased=True, fft=n > THRESHOLD_TO_USE_FFT, nlags=max_maxlag)[1:]
TypeError: acf() got an unexpected keyword argument 'unbiased'