April 2018
Beginner to intermediate
282 pages
6h 52m
English
Let's make some adjustments to the fit_predict method to include a decomposer in your pipeline, so that you can visualize high-dimensional data if necessary:
def fit_predict(self, X, y=None, scaler=True, decomposer={'name': PCA, 'args':[], 'kwargs': {'n_components': 2}}): """ fit_predict will train given estimator(s) and predict cluster membership for each sample """ shape = X.shape df_type = isinstance(X, pd.core.frame.DataFrame) if df_type: column_names = X.columns index = X.index if scaler == True: from sklearn.preprocessing import StandardScaler scaler = StandardScaler() X = scaler.fit_transform(X) if df_type: X = pd.DataFrame(X, index=index, columns=column_names) if decomposer ...