Search code examples
python-3.xscikit-learnsvmsentiment-analysissklearn-pandas

TypeError: __init__() got an unexpected keyword argument 'n_folds',sentiment_analysis_with_SVM

I am trying to implement svm for sentiment analysis, i trying to implement this gitlink https://github.com/jatinwarade/Sentiment-analysis-using-SVM/blob/master/SVM.ipynb.

from sklearn.model_selection import ShuffleSplit
from sklearn.model_selection import StratifiedKFold

i refered this as it says to change cross origin to model_selection since it is depricated Error: __init__() got an unexpected keyword argument 'n_splits' so I replaced with this

grid_svm = GridSearchCV(
    pipeline_svm, #object used to fit the data
    param_grid=param_svm, 
    refit=True,  # fit using all data, on the best detected classifier
    n_jobs=-1,  # number of cores to use for parallelization; -1 for "all cores" i.e. to run on all CPUs
    scoring='accuracy',#optimizing parameter
    cv=StratifiedKFold(liked_train,n_folds=5),
)

This Returns Error: 

TypeError                                 Traceback (most recent call last)
<ipython-input-49-61dd1e818fa4> in <module>
      5     n_jobs=-1,  # number of cores to use for parallelization; -1 for "all cores" i.e. to run on all CPUs
      6     scoring='accuracy',#optimizing parameter
----> 7     cv=StratifiedKFold(liked_train,n_folds=5),
      8 )

TypeError: __init__() got an unexpected keyword argument 'n_folds'

Please Help me solve this error

Solution

  • As you can see in the documentation for model_selected.StrafiedKFold, there is no keyword argument called n_folds and you should indeed use n_splits.

    Note, however, that the data should not be passed as an argument to the validator and by doing so, you're effectively passing liked_train as the argument for n_splits, which won't work. Rather, you should pass the data only to the fit of your grid_svm after initialization.