Uncaught exception? ValueError: SelectRates removed all features.

[Motorrat]

I have re-installed auto-sklearn 0.0.2 from the github and just want to share what I see in the log, It does not break the overall process.

/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/feature_selection/base.py:80: UserWarning: No features were selected: either the data is too noisy or the selection test too strict.
  UserWarning)
Process pynisher function call:
Traceback (most recent call last):
  File "/x/Redly/anaconda3/lib/python3.5/multiprocessing/process.py", line 249, in _bootstrap
    self.run()
  File "/x/Redly/anaconda3/lib/python3.5/multiprocessing/process.py", line 93, in run
    self._target(*self._args, **self._kwargs)
  File "/x/Redly/anaconda3/lib/python3.5/site-packages/pynisher/limit_function_call.py", line 83, in subprocess_func
    return_value = ((func(*args, **kwargs), 0))
  File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/evaluation/holdout_evaluator.py", line 148, in eval_holdout
    loss, opt_pred, valid_pred, test_pred = evaluator.fit_predict_and_loss()
  File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/evaluation/holdout_evaluator.py", line 59, in fit_predict_and_loss
    self.model.fit(X_train, Y_train)
  File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/base.py", line 62, in fit
    init_params=init_params)
  File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/classification.py", line 87, in pre_transform
    X, y, fit_params=fit_params, init_params=init_params)
  File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/base.py", line 131, in pre_transform
    X, fit_params = self.pipeline_._pre_transform(X, y, **fit_params)
  File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/pipeline.py", line 148, in _pre_transform
    .transform(Xt)
  File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/components/feature_preprocessing/select_rates.py", line 72, in transform
    "%s removed all features." % self.__class__.__name__)
ValueError: SelectRates removed all features.

[Motorrat]

n_components is too large: it will be set to 25
Process pynisher function call:
Traceback (most recent call last):
  File "/x/Redly/anaconda3/lib/python3.5/multiprocessing/process.py", line 249, in _bootstrap
    self.run()
  File "/x/Redly/anaconda3/lib/python3.5/multiprocessing/process.py", line 93, in run
    self._target(*self._args, **self._kwargs)
  File "/x/Redly/anaconda3/lib/python3.5/site-packages/pynisher/limit_function_call.py", line 83, in subprocess_func
    return_value = ((func(*args, **kwargs), 0))
  File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/evaluation/holdout_evaluator.py", line 148, in eval_holdout
    loss, opt_pred, valid_pred, test_pred = evaluator.fit_predict_and_loss()
  File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/evaluation/holdout_evaluator.py", line 59, in fit_predict_and_loss
    self.model.fit(X_train, Y_train)
  File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/base.py", line 62, in fit
    init_params=init_params)
  File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/classification.py", line 87, in pre_transform
    X, y, fit_params=fit_params, init_params=init_params)
  File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/base.py", line 131, in pre_transform
    X, fit_params = self.pipeline_._pre_transform(X, y, **fit_params)
  File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/pipeline.py", line 147, in _pre_transform
    Xt = transform.fit(Xt, y, **fit_params_steps[name]) \
  File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/components/feature_preprocessing/fast_ica.py", line 33, in fit
    self.preprocessor.fit(X)
  File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/decomposition/fastica_.py", line 523, in fit
    self._fit(X, compute_sources=False)
  File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/decomposition/fastica_.py", line 479, in _fit
    compute_sources=compute_sources, return_n_iter=True)
  File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/decomposition/fastica_.py", line 335, in fastica
    W, n_iter = _ica_par(X1, **kwargs)
  File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/decomposition/fastica_.py", line 116, in _ica_par
    warnings.warn('FastICA did not converge. Consider increasing '
UserWarning: FastICA did not converge. Consider increasing tolerance or the maximum number of iterations.

[Motorrat]

`n_components is too large: it will be set to 25 Process pynisher function call: Traceback (most recent call last): File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/components/feature_preprocessing/fastica.py", line 33, in fit self.preprocessor.fit(X) File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/decomposition/fastica.py", line 523, in fit self._fit(X, computesources=False) File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/decomposition/fastica.py", line 479, in _fit compute_sources=compute_sources, return_niter=True) File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/decomposition/fastica.py", line 335, in fastica W, n_iter = _icapar(X1, **kwargs) File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/decomposition/fastica.py", line 108, in _ica_par

• gwtx[:, np.newaxis] * W) File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/decomposition/fastica.py", line 55, in _sym_decorrelation s, u = linalg.eigh(np.dot(W, W.T)) File "/x/Redly/anaconda3/lib/python3.5/site-packages/scipy/linalg/decomp.py", line 288, in eigh a1 = _asarray_validated(a, check_finite=check_finite) File "/x/Redly/anaconda3/lib/python3.5/site-packages/scipy/_lib/_util.py", line 228, in _asarray_validated a = toarray(a) File "/x/Redly/anaconda3/lib/python3.5/site-packages/numpy/lib/function_base.py", line 1033, in asarray_chkfinite "array must not contain infs or NaNs") ValueError: array must not contain infs or NaNs

During handling of the above exception, another exception occurred:

Traceback (most recent call last): File "/x/Redly/anaconda3/lib/python3.5/multiprocessing/process.py", line 249, in _bootstrap self.run() File "/x/Redly/anaconda3/lib/python3.5/multiprocessing/process.py", line 93, in run self._target(_self._args, _self._kwargs) File "/x/Redly/anaconda3/lib/python3.5/site-packages/pynisher/limit_function_call.py", line 83, in subprocess_func return_value = ((func(_args, _kwargs), 0)) File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/evaluation/holdout_evaluator.py", line 148, in eval_holdout loss, opt_pred, valid_pred, test_pred = evaluator.fit_predict_and_loss() File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/evaluation/holdout_evaluator.py", line 59, in fit_predict_and_loss self.model.fit(X_train, Y_train) File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/base.py", line 62, in fit init_params=init_params) File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/classification.py", line 87, in pre_transform X, y, fit_params=fit_params, init_params=init_params) File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/base.py", line 131, in pre_transform X, fitparams = self.pipeline._pre_transform(X, y, _fit_params) File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/pipeline.py", line 147, in _pre_transform Xt = transform.fit(Xt, y, _fit_params_steps[name]) \ File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/components/feature_preprocessing/fast_ica.py", line 36, in fit raise ValueError("Bug in scikit-learn: https://github.com/scikit-learn/scikit-learn/pull/2738") ValueError: Bug in scikit-learn: https://github.com/scikit-learn/scikit-learn/pull/2738 `

[mfeurer]

Similar to #151, this output will no longer be there with this afternoons release.

[Motorrat]

I have re-installed the autosklearn 1.1 but still see the above error:

n_components is too large: it will be set to 23 Process pynisher function call: Traceback (most recent call last): File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/components/feature_preprocessing/fastica.py", line 33, in fit self.preprocessor.fit(X) File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/decomposition/fastica.py", line 523, in fit self._fit(X, computesources=False) File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/decomposition/fastica.py", line 479, in _fit compute_sources=compute_sources, return_niter=True) File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/decomposition/fastica.py", line 335, in fastica W, n_iter = _icapar(X1, **kwargs) File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/decomposition/fastica.py", line 108, in _ica_par

• gwtx[:, np.newaxis] * W) File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/decomposition/fastica.py", line 55, in _sym_decorrelation s, u = linalg.eigh(np.dot(W, W.T)) File "/x/Redly/anaconda3/lib/python3.5/site-packages/scipy/linalg/decomp.py", line 288, in eigh a1 = _asarray_validated(a, check_finite=check_finite) File "/x/Redly/anaconda3/lib/python3.5/site-packages/scipy/_lib/_util.py", line 228, in _asarray_validated a = toarray(a) File "/x/Redly/anaconda3/lib/python3.5/site-packages/numpy/lib/function_base.py", line 1033, in asarray_chkfinite "array must not contain infs or NaNs") ValueError: array must not contain infs or NaNs

During handling of the above exception, another exception occurred:

Traceback (most recent call last): File "/x/Redly/anaconda3/lib/python3.5/multiprocessing/process.py", line 249, in _bootstrap self.run() File "/x/Redly/anaconda3/lib/python3.5/multiprocessing/process.py", line 93, in run self._target(*self._args, *self._kwargs) File "/x/Redly/anaconda3/lib/python3.5/site-packages/pynisher/limit_function_call.py", line 83, in subprocess_func return_value = ((func(args, kwargs), 0)) File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/evaluation/holdout_evaluator.py", line 148, in eval_holdout loss, opt_pred, valid_pred, test_pred = evaluator.fit_predict_and_loss() File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/evaluation/holdout_evaluator.py", line 59, in fit_predict_and_loss self._fit_and_suppress_warnings(self.model, X_train, Y_train) File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/evaluation/abstract_evaluator.py", line 335, in _fit_and_suppress_warnings model = model.fit(X, y) File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/base.py", line 62, in fit init_params=init_params) File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/classification.py", line 87, in pre_transform X, y, fit_params=fit_params, init_params=init_params) File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/base.py", line 131, in pre_transform X, fitparams = self.pipeline._pre_transform(X, y, fit_params) File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/pipeline.py", line 147, in _pre_transform Xt = transform.fit(Xt, y, fit_params_steps[name]) \ File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/components/feature_preprocessing/fast_ica.py", line 36, in fit raise ValueError("Bug in scikit-learn: https://github.com/scikit-learn/scikit-learn/pull/2738") ValueError: Bug in scikit-learn: https://github.com/scikit-learn/scikit-learn/pull/2738 Process pynisher function call: Traceback (most recent call last): File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/components/feature_preprocessing/fastica.py", line 33, in fit self.preprocessor.fit(X) File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/decomposition/fastica.py", line 523, in fit self._fit(X, computesources=False) File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/decomposition/fastica.py", line 479, in _fit compute_sources=compute_sources, return_niter=True) File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/decomposition/fastica.py", line 335, in fastica W, n_iter = _ica_par(X1, kwargs) File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/decomposition/fastica_.py", line 108, in _ica_par

• gwtx[:, np.newaxis] * W) File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/decomposition/fastica.py", line 55, in _sym_decorrelation s, u = linalg.eigh(np.dot(W, W.T)) File "/x/Redly/anaconda3/lib/python3.5/site-packages/scipy/linalg/decomp.py", line 288, in eigh a1 = _asarray_validated(a, check_finite=check_finite) File "/x/Redly/anaconda3/lib/python3.5/site-packages/scipy/_lib/_util.py", line 228, in _asarray_validated a = toarray(a) File "/x/Redly/anaconda3/lib/python3.5/site-packages/numpy/lib/function_base.py", line 1033, in asarray_chkfinite "array must not contain infs or NaNs") ValueError: array must not contain infs or NaNs

During handling of the above exception, another exception occurred:

Traceback (most recent call last): File "/x/Redly/anaconda3/lib/python3.5/multiprocessing/process.py", line 249, in _bootstrap self.run() File "/x/Redly/anaconda3/lib/python3.5/multiprocessing/process.py", line 93, in run self._target(*self._args, *self._kwargs) File "/x/Redly/anaconda3/lib/python3.5/site-packages/pynisher/limit_function_call.py", line 83, in subprocess_func return_value = ((func(args, kwargs), 0)) File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/evaluation/holdout_evaluator.py", line 148, in eval_holdout loss, opt_pred, valid_pred, test_pred = evaluator.fit_predict_and_loss() File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/evaluation/holdout_evaluator.py", line 59, in fit_predict_and_loss self._fit_and_suppress_warnings(self.model, X_train, Y_train) File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/evaluation/abstract_evaluator.py", line 335, in _fit_and_suppress_warnings model = model.fit(X, y) File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/base.py", line 62, in fit init_params=init_params) File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/classification.py", line 87, in pre_transform X, y, fit_params=fit_params, init_params=init_params) File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/base.py", line 131, in pre_transform X, fitparams = self.pipeline._pre_transform(X, y, fit_params) File "/x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/pipeline.py", line 147, in _pre_transform Xt = transform.fit(Xt, y, **fit_params_steps[name]) \ File "/x/Redly/anaconda3/lib/python3.5/site-packages/autosklearn/pipeline/components/feature_preprocessing/fast_ica.py", line 36, in fit raise ValueError("Bug in scikit-learn: https://github.com/scikit-learn/scikit-learn/pull/2738") ValueError: Bug in scikit-learn: https://github.com/scikit-learn/scikit-learn/pull/2738 n_components is too large: it will be set to 23

[Motorrat]

and some time later in the log this has appeared: /x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/linear_model/base.py:284: RuntimeWarning: overflow encountered in exp np.exp(prob, prob) /x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/discriminant_analysis.py:688: UserWarning: Variables are collinear warnings.warn("Variables are collinear") /x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/discriminantanalysis.py:688: UserWarning: Variables are collinear warnings.warn("Variables are collinear") /x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/decomposition/fastica.py:116: UserWarning: FastICA did not converge. Consider increasing tolerance or the maximum number of iterations. warnings.warn('FastICA did not converge. Consider increasing ' /x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/discriminant_analysis.py:688: UserWarning: Variables are collinear warnings.warn("Variables are collinear") /x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/discriminant_analysis.py:688: UserWarning: Variables are collinear warnings.warn("Variables are collinear") /x/Redly/anaconda3/lib/python3.5/site-packages/sklearn/linear_model/base.py:284: RuntimeWarning: overflow encountered in exp np.exp(prob, prob)

[mfeurer]

Is this only in the log file or also in the command line output? I moved all warnings into the log files, but the output is still there.

Maybe there should be two log files, one with debug output and one with info output that is shows the progress of auto-sklearn.

[lsorber]

Still seeing ValueError: SelectRates removed all features. during training on 0.1.2.

[mfeurer]

Thanks for pointing that out.

[lsorber]

Also seeing ValueError: Numerical problems in QDA. QDA.scalings_ contains values <= 0.0, but it's less common (on an np.float64 feature matrix where all features are in [0, 1]).

[mfeurer]

This failures should no longer be visible on stdout/stderr in the latest release (0.2.0).