Add Multiclass One-Vs-Rest Classifier

[tezzytezzy]

Ideally One-Vs-Rest classifier to be added to an arsenal of binary classifications.

Using the already used base logistic Regression model as input as below,

OVR = OneVsRest(classifier=LR)
OVR_paramGrid = (ParamGridBuilder().build())
classification_model_and_paramgrid_list.append([OVR, OVR_paramGrid])

the line cvModel = cv.fit(df_train) throws the following exception.

IllegalArgumentException: 'Field "rawPrediction" does not exist.\nAvailable fields: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 1_INDEX, 1_VEC, 3_INDEX, 3_VEC, 4_INDEX, 4_VEC, 6_INDEX, 6_VEC, 7_INDEX, 7_VEC, 9_INDEX, 9_VEC, 10_INDEX, 10_VEC, 12_INDEX, 12_VEC, 14_INDEX, 14_VEC, 15_INDEX, 15_VEC, 17_INDEX, 17_VEC, 19_INDEX, 19_VEC, 20_INDEX, 20_VEC, label, features, std_features, pca_features, CrossValidator_01aa8f348031_rand, prediction'