Using RandomForestClassificatier on vectors for predicting labels gives "Found input variables with inconsistent numbers of samples"

[berserker1]

I am learning from Active Regression tutorial page but it has not taken up the case of applying learners to more than one dimension vectors ( I was not able to find a specific example in the doc for this, so please point if you know one ).

In the function named my_stuff

My learner is

regressor = ActiveLearner(
        estimator=RandomForestClassifier(),
        query_strategy=entropy_sampling,
        X_training=X_training, y_training=y_training.ravel()
    )

My dataset X is (13084, 50) ( meaning 13084 vectors each having 50 length ) and y is (13084, 1) ( similar meaning ).

Here X_training is (5, 50) and y_training is (5, 1).
In this section of the code( taken blatantly from the tutorial page mentioned above ):

for idx in range(n_queries):
        query_idx, query_instance = regressor.query(X)
        print(query_idx, 'query_idx', X_training.shape, y_training.shape)
        regressor.teach(X[query_idx].reshape(-1, 1), y[query_idx].reshape(-1, 1))

The program ended abruptly, so upon using python debugger I found the error:

ValueError: Found input variables with inconsistent numbers of samples: [50, 1]
> /path/to/file/predict.py(286)my_stuff()
-> regressor.teach(X[query_idx].reshape(-1, 1), y[query_idx].reshape(-1, 1))

regressor
Here X[query_idx].reshape(-1, 1) has shape (50, 1) and y[query_idx].reshape(-1, 1) has shape (1, 1).

What would be the correct procedure for the teach procedure?

[BoyanH]

You should perhaps use regressor.teach(X[query_idx].reshape(1, -1), y[query_idx].reshape(1, -1)) (Notice (1, -1) instead of (-1,1)). The idea is that X[query_idx] returns just a single instance, but you need a whole batch in regressor.teach so that converts, in your case, a 50-dimensional vector to 1x50 matrix. The other way around ((-1, 1)) makes no sense, since you are effectively turning each of your 50 features in training instances, each a 1-dimensional vector.

[berserker1]

So you are saying that I should give one (1, 50) matrix and another (1, 1) matrix? I have tried that also and that yeilded this error:

ValueError: the dimensions of the new training data and label mustagree with the training data and labels provided so far
> /file//new_work/JointBERT/predict.py(286)my_stuff()
-> regressor.teach(X[query_idx].reshape(1, -1), y[query_idx].reshape(1, -1))

X[query_idx].reshape(1, -1) has shape (1, 50) y[query_idx].reshape(1, -1) has shape (1, 1)

[BoyanH]

You also need to make sure the training data is the correct shape, as you expect it according to your classification problem. Both the X_training and y_training arguments when creating the learner, as well as the corresponding arguments to regressor.teach should be of the same shape. They are both essentially using the fit() method of sklearn.

I think chaging y_training=y_training.ravel() to y_training=y_training should solve the problem in your case, but you also need to make sure your data is really encoded the way you expect.

[berserker1]

Thanks for help, I will look into it further but for now I will try to give the current scenario:

• X_training shape is (5, 50), my n_initial is 5.
• y_training shape is (5, 1), it gave a warning that Randomforest expects a 1d array that is why I did y_training.ravel() to give a (5,) shape vector as input

Rest shapes of the matrices I have explained above.

[BoyanH]

So after ravel, your training labels have shape (5,), so I think you need to to use regressor.teach(X[query_idx].reshape(1, -1), y[query_idx].reshape(1)), which will reshape y[query_idx] into (1,).

[berserker1]

Hello, so I tried this and it worked!, although I used y[query_idx].reshape(1, -1) because in the tutorial link a vector of shape (1, x) was passed ( as per my understanding ).

I think a small tut on multi dimensional vectors will do good but I think even this issue will be work as a nice reference for now. I am closing this issue for now