Browse Source

grond docs: update weighting plot, reviewed noise analyser doc strings

priors
Henriette Sudhaus 2 years ago
parent
commit
a9b5da003d
3 changed files with 2199 additions and 906 deletions
  1. +1072
    -900
      docs/source/images/illu_combi_weights.svg
  2. +1120
    -0
      docs/source/images/illu_combi_weights_future.svg
  3. +7
    -6
      src/analysers/noise_analyser/analyser.py

+ 1072
- 900
docs/source/images/illu_combi_weights.svg
File diff suppressed because it is too large
View File


+ 1120
- 0
docs/source/images/illu_combi_weights_future.svg
File diff suppressed because it is too large
View File


+ 7
- 6
src/analysers/noise_analyser/analyser.py View File

@ -160,10 +160,11 @@ class NoiseAnalyser(Analyser):
within a 30 min time window before the start of the set pre-event noise
window, only a warning is thrown.
It is also possible to remove data where the noise level exceeds the median
by a given factor (by setting there weight to 0). This can be done
exclusively (``mode='weeding'``) or in combination with weighting
(``mode='weighting'``).
It is further possible to disregard data with a noise level exceeding the
median by a given ``cutoff`` factor. These weights are set to 0. This can be
done exclusively (``mode='weeding'``) such that noise weights are either
1 or 0, or in combination with weighting below the median-times-cutoff
noise level (``mode='weighting'``).
'''
def __init__(self, nwindows, pre_event_noise_duration,
@ -311,8 +312,8 @@ class NoiseAnalyserConfig(AnalyserConfig):
cutoff = Float.T(
optional=True,
help='Set weight to zero, when noise level exceeds median by given '
'factor.')
help='Set weight to zero, when noise level exceeds median by the '
'given cutoff factor.')
def get_analyser(self):
return NoiseAnalyser(


Loading…
Cancel
Save