Refitting a model using the training data

I am working with a group who will generate a model with Sklearn (usually a logistic regression) using training and tes开发者_开发技巧t data, and then, after the model is created, refit the model using a combination of the training and test data. Is this a legitimate thing to do?