Trying out Python package tsfresh I run into issues in the first steps. Given a series how to (automatically) make features for it? This snippet produces different errors based on which part I try.
import tsfresh
import pandas as pd
import numpy as np
#tfX, tfy = tsfresh.utilities.dataframe_functions.make_forecasting_frame(pd.Series(np.random.randn(1000)/50), kind='float64', max_timeshift=50, rolling_direction=1)
#rf = tsfresh.extract_relevant_features(tfX, y=tfy, n_jobs=1, column_id='id')
tfX, tfy = tsfresh.utilities.dataframe_functions.make_forecasting_frame(pd.Series(np.random.randn(1000)/50), kind=1, max_timeshift=50, rolling_direction=1)
rf = tsfresh.extract_relevant_features(tfX, y=tfy, n_jobs=1, column_id='id')
The errors are in the first case
""" Traceback (most recent call last): File "C:\Users\user\Anaconda3\envs\env1\lib\multiprocessing\pool.py", line 119, in worker result = (True, func(*args, **kwds)) File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\tsfresh\utilities\distribution.py", line 38, in _function_with_partly_reduce results = list(itertools.chain.from_iterable(results)) File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\tsfresh\utilities\distribution.py", line 37, in results = (map_function(chunk, **kwargs) for chunk in chunk_list) File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\tsfresh\feature_extraction\extraction.py", line 358, in _do_extraction_on_chunk return list(_f()) File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\tsfresh\feature_extraction\extraction.py", line 350, in _f result = [("", func(data))] File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\tsfresh\feature_extraction\feature_calculators.py", line 193, in variance_larger_than_standard_deviation y = np.var(x) File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\numpy\core\fromnumeric.py", line 3157, in var **kwargs) File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\numpy\core_methods.py", line 110, in _var arrmean, rcount, out=arrmean, casting='unsafe', subok=False) TypeError: unsupported operand type(s) for /: 'str' and 'int' """
and in the second case
""" Traceback (most recent call last): File "C:\Users\user\Anaconda3\envs\env1\lib\multiprocessing\pool.py", line 119, in worker result = (True, func(*args, **kwds)) File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\tsfresh\utilities\distribution.py", line 38, in _function_with_partly_reduce results = list(itertools.chain.from_iterable(results)) File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\tsfresh\utilities\distribution.py", line 37, in results = (map_function(chunk, **kwargs) for chunk in chunk_list) File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\tsfresh\feature_extraction\extraction.py", line 358, in _do_extraction_on_chunk return list(_f()) File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\tsfresh\feature_extraction\extraction.py", line 345, in _f result = func(data, param=parameter_list) File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\tsfresh\feature_extraction\feature_calculators.py", line 1752, in friedrich_coefficients coeff = _estimate_friedrich_coefficients(x, m, r) File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\tsfresh\feature_extraction\feature_calculators.py", line 145, in _estimate_friedrich_coefficients result.dropna(inplace=True) File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\pandas\core\frame.py", line 4598, in dropna result = self.loc(axis=axis)[mask] File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\pandas\core\indexing.py", line 1500, in getitem return self._getitem_axis(maybe_callable, axis=axis) File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\pandas\core\indexing.py", line 1859, in _getitem_axis if is_iterator(key): File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\pandas\core\dtypes\inference.py", line 157, in is_iterator return hasattr(obj, 'next') File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\pandas\core\generic.py", line 5065, in getattr if self._info_axis._can_hold_identifiers_and_holds_name(name): File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\pandas\core\indexes\base.py", line 3984, in _can_hold_identifiers_and_holds_name return name in self File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\pandas\core\indexes\category.py", line 327, in contains return contains(self, key, container=self._engine) File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\pandas\core\arrays\categorical.py", line 188, in contains loc = cat.categories.get_loc(key) File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\pandas\core\indexes\interval.py", line 770, in get_loc start, stop = self._find_non_overlapping_monotonic_bounds(key) File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\pandas\core\indexes\interval.py", line 717, in _find_non_overlapping_monotonic_bounds start = self._searchsorted_monotonic(key, 'left') File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\pandas\core\indexes\interval.py", line 681, in _searchsorted_monotonic return sub_idx._searchsorted_monotonic(label, side) File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\pandas\core\indexes\base.py", line 4755, in _searchsorted_monotonic return self.searchsorted(label, side=side) File "C:\Users\user\Anaconda3\envs\env1\lib\site-packages\pandas\core\base.py", line 1501, in searchsorted return self._values.searchsorted(value, side=side, sorter=sorter) TypeError: Cannot cast array data from dtype('float64') to dtype('
np.version, tsfresh.version are ('1.15.4', 'unknown'). I installed tsfresh using conda, probably from conda-forge. I am on Windows 10. Using another kernel with np.version, tsfresh.version ('1.15.4', '0.11.2') lead to the same results.
Trying the first couple of cells from timeseries_forecasting_basic_example.ipynb yields the casting error as well.
