Replace to_dataframe() by new stats export methods

[gplssm]

• Find all usages of to_dataframe()
• Replace by according new stats export function and adjust dependent code
• Make sure to_dataframe() does not contain any figures that are not included in the new functions and delete to_dataframe()

[miguelparadacontzen]

Like we spoke, better not delete to_dataframe() because it gives information (a list of nodes and edges with some attributes) that is different than that of the stats functions and might be useful. in particular, we use it for the regression test function, in order to compare if two networks are equal (if they have the same nodes and edges with the same attributes)