python – ValueError: Unknown label type: continuous
python – ValueError: Unknown label type: continuous
The solution of your problem is that you need regression model instead of classification model so:
istead of these two lines:
from sklearn.svm import SVC
..
..
models.append((SVM, SVC()))
use these:
from sklearn.svm import SVR
..
..
models.append((SVM, SVR()))
Classifier expect in Y_train only integer values (labels of classes). But it get float and raise this error.
If you doing a regression, use the regressors instead of classifiers.
Or if you need a classification, check y_train. May be this part of your code transform it to float:
imp = Imputer(missing_values=NaN, strategy=median, axis=0)
array = df.values
imp.fit(array)
array_imp = imp.transform(array)
Y = array_imp[:,4]
try change it to
Y = array[:,4] # take it from not changed data
imp = Imputer(missing_values=NaN, strategy=median, axis=0)
array = df.values
imp.fit(array)
array_imp = imp.transform(array)