Issues with auto-derive xfac from manual metadata

[veenstrajelmer]

Issues:
xfac=True (empirisch) according to compfile, but xfac=False would be better according to AB/KD. Accidentally already added to stats_xfac0 in example files. Prevent mixing of xfac=False components with xfac=True prediction:

• L9PFM
• Q1

Empirisch/theoretisch is not mentioned in component file (so assuming xfac=True), but xfac=False resulted in good reproductions of components/prediction. Indeed theoretical f/u were used according to KD:

• AUKFPFM
• CUXHVN
• K13APFM
• NORTHCMRT (although "theor. f" for SA/SM)

This can be tested with xfac = COMP_merged.xfac in predictie_2019_frommergedcomp.py. If conclusive, update stats_xfac0 in example scripts and maybe update excel file. Comparing metadata is possible with:

import os
import hatyan

dir_testdata = 'C:\\DATA\\hatyan_data_acceptancetests'
stats_xfac0 = ['A12','ABDN','AUKFPFM','BOURNMH','CROMR','CUXHVN','D15','DEVPT','DOVR','F16','F3PFM','FELSWE','FISHGD','IMMHM','J6','K13APFM','K14PFM','KINLBVE','L9PFM','LEITH','LLANDNO','LOWST','NEWHVN','NEWLN','NORTHCMRT','NORTHSS','PORTSMH','Q1','SHEERNS','STORNWY','WEYMH','WHITBY','WICK']
selected_stations = ["L9PFM","Q1","AUKFPFM","CUXHVN","K13APFM","NORTHCMRT"]

for current_station in selected_stations:
    print(f'current_station: {current_station}')
    if current_station in stats_xfac0:
        xfac=False
    else:
        xfac=True
    file_comp1 = os.path.join(dir_testdata,'predictie2019',f'{current_station}_ana.txt')
    comp_fromfile = hatyan.read_components(filename=file_comp1)
    compfile_xfac = comp_fromfile.attrs["xfac"]
    print(f'{current_station}: xfac={compfile_xfac} in file, xfac={xfac} in practice')
    # assert compfile_xfac == xfac

After fix, also re-enable in acceptance tests again:

• After implementing 139 improve metadata in functions and component file #218, it is not possible anymore to make predictions with different xfac than derived from components file, so these stations cannot be reproduced anymore.
• this is all acceptable, but useful to update the data files to fix all example testcases again.
• CUXHVN in numbering_extremes.py
• AUKFPFM in predictie_2019_frommergedcomp.py