Dealing with missing data

NLeSC / mcfly

A deep learning tool for time series classification and regression

Apache License 2.0

362 stars 82 forks source link

Dealing with missing data #260

Open machielg opened 4 years ago

machielg commented 4 years ago

Not sure if it’s a feature request or something for the documentation but I am wondering how to deal with missing data. I have for each user a multi variate time series but not always complete or of equal size/length. Should I just leave NaNs or use interpolation? In the second case that could be feature of mcfly.

cwmeijer commented 4 years ago

Mcfly doesn't have any options for imputing missing values, but there are many available in sklearn (see https://scikit-learn.org/stable/modules/impute.html). We don't plan to implement any at the moment, but I will leave this issue open until our team has made a decision on the matter.