python - What is this error with sklearn's skpca.fit when doing a PCA analysis

Question

I am doing a simple PCA analysis with some satellite data. All the land points are removed, the mean and standard deviation are close to 0 and 1. However I get

from sklearn import preprocessing
scaler  = preprocessing.StandardScaler()
scaler_sst = scaler.fit(sss_data)

import joblib
joblib.dump(scaler_sst, './scaler_sst.pkl', compress=9)
scaler_sst = joblib.load('./scaler_sst.pkl')

X = scaler_sst.transform(sss_data)

print(X.mean())
print(X.std())
#X.shape

5.7725416769826885e-15
0.9999999999999993

from sklearn.decomposition import pca
skpca=pca.PCA()
skpca.fit(X)

---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
~/miniconda3/envs/py3_std_maps/lib/python3.8/site-packages/IPython/core/formatters.py in __call__(self, obj, include, exclude)
    968 
    969             if method is not None:
--> 970                 return method(include=include, exclude=exclude)
    971             return None
    972         else:

~/miniconda3/envs/py3_std_maps/lib/python3.8/site-packages/sklearn/base.py in _repr_mimebundle_(self, **kwargs)
    461     def _repr_mimebundle_(self, **kwargs):
    462         """Mime bundle used by jupyter kernels to display estimator"""
--> 463         output = {"text/plain": repr(self)}
    464         if get_config()["display"] == 'diagram':
    465             output["text/html"] = estimator_html_repr(self)

~/miniconda3/envs/py3_std_maps/lib/python3.8/site-packages/sklearn/base.py in __repr__(self, N_CHAR_MAX)
    273 
    274         # use ellipsis for sequences with a lot of elements
--> 275         pp = _EstimatorPrettyPrinter(
    276             compact=True, indent=1, indent_at_name=True,
    277             n_max_elements_to_show=N_MAX_ELEMENTS_TO_SHOW)

~/miniconda3/envs/py3_std_maps/lib/python3.8/site-packages/sklearn/utils/_pprint.py in __init__(self, indent, width, depth, stream, compact, indent_at_name, n_max_elements_to_show)
    162         if self._indent_at_name:
    163             self._indent_per_level = 1  # ignore indent param
--> 164         self._changed_only = get_config()['print_changed_only']
    165         # Max number of elements in a list, dict, tuple until we start using
    166         # ellipsis. This also affects the number of arguments of an estimators

KeyError: 'print_changed_only'
---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
~/miniconda3/envs/py3_std_maps/lib/python3.8/site-packages/IPython/core/formatters.py in __call__(self, obj)
    700                 type_pprinters=self.type_printers,
    701                 deferred_pprinters=self.deferred_printers)
--> 702             printer.pretty(obj)
    703             printer.flush()
    704             return stream.getvalue()

~/miniconda3/envs/py3_std_maps/lib/python3.8/site-packages/IPython/lib/pretty.py in pretty(self, obj)
    392                         if cls is not object 
    393                                 and callable(cls.__dict__.get('__repr__')):
--> 394                             return _repr_pprint(obj, self, cycle)
    395 
    396             return _default_pprint(obj, self, cycle)

~/miniconda3/envs/py3_std_maps/lib/python3.8/site-packages/IPython/lib/pretty.py in _repr_pprint(obj, p, cycle)
    698     """A pprint that just redirects to the normal repr function."""
    699     # Find newlines and replace them with p.break_()
--> 700     output = repr(obj)
    701     lines = output.splitlines()
    702     with p.group():

~/miniconda3/envs/py3_std_maps/lib/python3.8/site-packages/sklearn/base.py in __repr__(self, N_CHAR_MAX)
    273 
    274         # use ellipsis for sequences with a lot of elements
--> 275         pp = _EstimatorPrettyPrinter(
    276             compact=True, indent=1, indent_at_name=True,
    277             n_max_elements_to_show=N_MAX_ELEMENTS_TO_SHOW)

~/miniconda3/envs/py3_std_maps/lib/python3.8/site-packages/sklearn/utils/_pprint.py in __init__(self, indent, width, depth, stream, compact, indent_at_name, n_max_elements_to_show)
    162         if self._indent_at_name:
    163             self._indent_per_level = 1  # ignore indent param
--> 164         self._changed_only = get_config()['print_changed_only']
    165         # Max number of elements in a list, dict, tuple until we start using
    166         # ellipsis. This also affects the number of arguments of an estimators

KeyError: 'print_changed_only'

The error occurs at the part skpca.fit(X). I reinstalled the sklearn package and scikit packages. I have used a PCA analysis before with sklearn and this has never happened.

question from:https://stackoverflow.com/questions/65924425/what-is-this-error-with-sklearns-skpca-fit-when-doing-a-pca-analysis

Answer 1

I don't know the answer but maybe this is a bug in sklearn: Try:

import sklearn
sklearn.get_config()

In my case it returns a dict:

{'assume_finite': False, 'working_memory': 1024, 'print_changed_only': False}

The error indicates that the print_changend_only does not exist for you. My sklearn version is '0.21.2' on python 3.6. Maybe it helps to downgrade the sklearn version?