Feature 2003 briercl#2034
Merged
JohnHalleyGotway merged 28 commits intodevelopfrom feature_2003_brierclFeb 2, 2022
+519-123
Commits
Commits on Jan 20, 2022
Commits on Jan 25, 2022
Commits on Jan 27, 2022
Commits on Jan 28, 2022
- committed
Per #2003, this gets BRIERCL and BSS to appear in the Series-Analysis output. But I have 2 concerns... doing a deep copy of cdf_info for each grid point seems like a lot of wasted space. Consider c…
committedPer #2003, replace PairBase::climo_cdf from a ClimoCDFInfo object with a PairBase::climo_cdf_ptr pointer to one. This is to needed avoid creating separate ClimoCDFInfo objects for each grid point i…
committed- committed
Per #2003, update the Ensemble-Stat, Grid-Stat, Point-Stat, and Stat-Analysis application code to hande the ClimoCDFInfo pointer.
committedPer #2003, working a bit backwards here. Updating the Series-Analysis config files to indicate the logic I'm adding to the tool. If block_size <= 0, automatically set it the full dimension of the v…
committed
Commits on Jan 29, 2022
Commits on Jan 31, 2022
- committed
- committed
Per #2003, ci-run-unit, when aggregating MPR or ORANK lines in Stat-Analysis, need to store the aggregation object in the map BEFORE storing the pointer to the CDF thresh array. The opposite doesn'…
committed- authored
- committed
Per #2003, correct logic for letting block_size = grid dimension. Can't allocate PairDataPoint objects until the grid is actually defined.
committed- committed
Per #2003, tighten up the logic in Series-Analysis for allocating PairDataPoint objects. ci-run-unit
committedPer #2003, update the block_size in SeriesAnalysisConfig_climo to set it to 0. That way it'll automatically resize to the dimension of the grid. Doing this make it run about 20 seconds faster, whic…
committedPer #2003, a little config file cleanup. Remove climo_cdf.write_bins from Series-Analysis config files since it has no impact on the output. That only applies to Grid-Stat and Point-Stat. But add e…
committedPer #2003, this change is unrelated but was requested by John Opatz. Tweaking a PB2NC log message to clarify that the reported time range is the input observation timestamps.
committed