Define better X_train and X_test

[terry07]

You may need to elaborate the readme file. During the introduction of the two methods, you use the term X_test for describing the unknown instances that you want to predict, but your documentation fluctuates at this point:

--> y_pred = label_propagation(y_labeled, X_labeled, X_unlabeled, X_train, gaussian_kernel(std), mu) Thus, in the above line you should replace the X_train with X_test variable, as well as into the code.

Thanks for the implementation.

[RobRomijnders]

Hi @terry07,

Thanks for diving into this issue!

It seems you have a clear understanding of the code (it has been a while for me). Might you want to submit a pull request? Then future people will benefit from your insights.

Thanks in advance, Rob