Feauture 1809 gen_prob#1966
Closed
JohnHalleyGotway wants to merge 25 commits intodevelopfrom feauture_1809_gen_prob
+2,007-243
Commits
Commits on Nov 5, 2021
Per #1809, define the ProbGenInfo class and update the Makefile to compile it. Define ATCF offsets specific to ATCF EDECK GN lines. Update the is_match() logic slightly by moving the checking for c…
committedPer #1809, update the ProbInfoArray class to store arrays of both ProbRIRWInfo and ProbGenInfo objects.
committedPer #1809, since tc_pairs processes probabilities of RIRW and not genesis, update it to only process RIRW probabilities from the ProbInfoArray class rather than all of them.
committedPer #1809, this is definitely a work in progress. TC-Gen parses the genesis probability forecasts and now needs to be enhanced to actually verify them.
committedPer #1809, store the distance to land in the ProbInfoBase class and update the other classes accordingly to set it. Also, update tc_pairs to only process ATCF EDECK probability of RI lines.
committed- committed
Commits on Nov 11, 2021
Per #1809, add ProbInfoArray::n_technique() member function to count up the unique model names from the input.
committedPer #1809, update PCTInfo to add the set_fthresh() function and expand probability thresholds, as needed.
committedPer #1809, update TC-Gen to add options for probabilistic outputs when evaluating the genesis probability forecasts.
committedPer #1809, switch to consistently naming classes related to the probability of genesis as ProbGen rather than mixing it with GenProb, for consistency. Also define ProbGenPCTInfo class for storing P…
committed
Commits on Nov 12, 2021
Per #1809, saving off the current state of development before proceeding with changes to simplify the matching logic.
committedPer #1809, keep track of the min/max BEST track genesis times, and we only need 1 PCT table, not two.
committedPer #1809, make access to ProbGenInfoArray elements non-constant so that the tc-gen application can modify their contents.
committed- committed
Per #1809, this is a good stopping point. I have tc-gen accurately verifying genesis probabilities. Still need to add a probgen_mpr line type, a unit test, and documentaiton.
committed- committed
Commits on Nov 13, 2021
- committed
- committed
Per #1809, add GenesisLead to the ProbGenInfo class to track the lead time for which genesis is forecast.
committedPer #1809, update tc_gen to write PROB_GENESIS matched pair output to the newly-updated GENMPR line type.
committed- committed
Per #1809, use ProbGenPCTInfo to store pointers to the individual matched pair objects. This is cleaner than the last kludgy solution.
committed- committed