lmizrahi / etas Goto Github PK

Thanks for providing these codes, it is very useful. I have a demand that using ETAS to decluster, as you do in your paper (Mizrahi et al., 2021, SRL). Could you release this code or give some recommendations?

A few things could be cleaned up/homogenized throughout the repo:

• in inversion.py: naming of "auxiliary_start", "timewindow_start", etc. could be more intuitively named.
• in evaluate.py: check the docstring for the likelihood calculation class
• in inversion.py and evaluate.py: both files define a to_days function, but they are different. hopefully only one is needed and can be used for both.
• update the README

is not working because of division by 0.

• mu can't be fixed currently (because it is estimated independently of the other parameters)
• parameters cannot be fixed to values that are outside the range
• when a parameter (such as log10(mu)) takes infinite values, the calc_diff_to_before function doesn't work

Why can't my python3.8 install the etas package? What is the name of the specific package? Thanks!