Meaning of partial_fit method of scikit-learn

Question

I am curious about whether running the partial_fit method of scikit-learn corresponds to one iteration, one step, or one epoch? I am confused.

Answer 1

It runs a single epoch according to the documentation.

Perform one epoch of stochastic gradient descent on given samples.

This is only available for some training methods, like stochastic gradient descent (SGD), which use multiple passes over the training data (epochs).