Features that are read from a csv are cast to floats. However, certain measurement apparatus report "" if - for example - a specific element is not present. This means that the user has to do preprocessing to change these values to 0.
We could help the user by casting '' to float by setting these values to 0. We should still warn the user that this is happening, in case they do not realisze their data has missing values.
In the future, we could also give the user an option: a) set these values to 0, or b) drop these instances.
Features that are read from a csv are cast to floats. However, certain measurement apparatus report "" if - for example - a specific element is not present. This means that the user has to do preprocessing to change these values to 0.
We could help the user by casting '' to float by setting these values to 0. We should still warn the user that this is happening, in case they do not realisze their data has missing values.
In the future, we could also give the user an option: a) set these values to 0, or b) drop these instances.