fixed append for special parameters + channel mean

legend-exp / legend-data-monitor

LEGEND data monitoring

https://legend-data-monitor.readthedocs.io

MIT License

0 stars 5 forks source link

fixed append for special parameters + channel mean #112

Closed sofia-calgaro closed 1 year ago

sofia-calgaro commented 1 year ago

Previously, I hadn't tested the "saving": "append" case for special parameters too. The point is that we need to load also columns for the parameters we used to get the special one (eg baseline and wf_max for wf_max_rel). Of course this applies if the special parameter is based on not null parameters (see settings/special-parameters.json for more info).

sofia-calgaro commented 1 year ago

before merging, still need to

see if removing the parameters->parameter thing in utils.py gives problem or not (leaving it causes problems when using plotting.py functions - but maybe is related only to special parameters? re-check everythin)
check some ??? and !!! left in utils.py, analysis_data.py

sofia-calgaro commented 1 year ago

Checks for correctness of append option done for:

1 par, both special (wf_max_rel) oand not-special (baseline)
2 pars (baseline vs cuspEmax)
2 pars (baseline vs wf_max_rel)

sofia-calgaro commented 1 year ago

The evaluation of the mean column is fixed when using the append option. Before, multiple values were present for a fixed detector because doing old_df[param_mean] = new_df[param_mean] I was not considering that the two dataframes actually have different lengths. Now we map the new mean values saved in new_df into old_df, making the mapping through the channel ID, and then we merge horizontally the two dataframes.