Closed dominikbach closed 6 months ago
I agree with the concept that "sts should be provided for all internally called functions that produce output, to offer a systematic way to handle problems without interrupting batch processing".
The reason why I conceptualise "dynamic output" is to minimise the modifications of how these functions are called.
If we change all functions to [ sts, Y ] = function [ X ]
it will be a heavy update.
For example glm = pspm_glm [ X ]
will need to be updated into [sts, glm] = pspm_glm [ X ]
. As the result, all the functions which call pspm_glm
will need to be changed.
However, I admit, static outputs are theoretically more robust and easier to debug, and following the traditional coding style of PsPM. Therefore, I don't have a strong opinion to use either static or dynamic outputs. The only thing I think that is useful is to make the logic consistent between functions, which is what I was normally trying to improve.
Feature Description