You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Rewritten on local branch ISSUE843 (commits d2ab600a, 1474e02d).
Checked every name on the page against hera/measurements/meteorology/lowfreqdata/. Four things were wrong:
plotSeasonalHourly does not exist anywhere in the package.
calcHourlyDist takes Field (capital F) and normalization, not field and density.
plotProbContourf_bySeason takes plotField, a single column name, not fields=[...].
dateLinePlot, plotProbContourf and resampleSecondMoments were not documented at all.
The page now follows the structure of the sibling MeteoHighFreq page: data source format and docType, initialisation, all three analysis methods (with the exact columns addDatesColumns adds and the three valid normalization values), and all four plotting methods grouped by the dailyPlots / seasonalPlots accessor they hang off. It also records that the toolkit constant MeteoLowFreq and the database toolkit name lowFreqMeteorology differ, which the old page did not mention.
hera/tests/unit/test_lowfreq_docs_match_code.py parses the Python examples out of the page and resolves every lf.<accessor>.<method>(...) call and keyword argument against the real classes, so this cannot drift again silently. Against the old page it fails with exactly the four errors above.
Separate commit, unrelated to the docs: presenation.__init__, Plots.__init__, SeasonalPlots.__init__ and DailyPlots.__init__ each printed emoji debug lines, so every getToolkit(METEOROLOGY_LOWFREQ) wrote seven lines to stdout. Removed.
Activity
ilayfalach commentedon Sep 15, 2026
Rewritten on local branch
ISSUE843(commitsd2ab600a,1474e02d).Checked every name on the page against
hera/measurements/meteorology/lowfreqdata/. Four things were wrong:plotSeasonalHourlydoes not exist anywhere in the package.calcHourlyDisttakesField(capital F) andnormalization, notfieldanddensity.plotProbContourf_bySeasontakesplotField, a single column name, notfields=[...].dateLinePlot,plotProbContourfandresampleSecondMomentswere not documented at all.The page now follows the structure of the sibling MeteoHighFreq page: data source format and
docType, initialisation, all three analysis methods (with the exact columnsaddDatesColumnsadds and the three validnormalizationvalues), and all four plotting methods grouped by thedailyPlots/seasonalPlotsaccessor they hang off. It also records that the toolkit constantMeteoLowFreqand the database toolkit namelowFreqMeteorologydiffer, which the old page did not mention.hera/tests/unit/test_lowfreq_docs_match_code.pyparses the Python examples out of the page and resolves everylf.<accessor>.<method>(...)call and keyword argument against the real classes, so this cannot drift again silently. Against the old page it fails with exactly the four errors above.Separate commit, unrelated to the docs:
presenation.__init__,Plots.__init__,SeasonalPlots.__init__andDailyPlots.__init__each printed emoji debug lines, so everygetToolkit(METEOROLOGY_LOWFREQ)wrote seven lines to stdout. Removed.ilayfalach commentedon Sep 15, 2026
PR #1078 opened.
Merge pull request #1078 from KaplanOpenSource/ISSUE843