Decreased performance of calibration execution time
Decreased performance of calibration execution time
Since merge of branch 132-runmodel_lag-add-warmupqsim-to-outputsmodel into dev the calibration schedule test lasts more than 60 minutes to run although it lasts less than 25 minutes before.
Activity
- Dorchies David changed milestone to %v1.7.0
changed milestone to %v1.7.0
- Dorchies David added BUG label
added BUG label
- Dorchies David assigned to @david.dorchies
assigned to @david.dorchies
- Dorchies David created merge request !57 (merged) to address this issue
created merge request !57 (merged) to address this issue
- Dorchies David mentioned in merge request !57 (merged)
mentioned in merge request !57 (merged)
- Author Developer
Here's a exploration of which model are in trouble by benchmarking there execution time in calibration configuration:
model 1.6.12 1.6.12.9000 evolution GR1A 0.08558205 0.11381404 0.329882142341765 GR2M 0.21442601 0.21779505 0.0157118998763256 GR4J 2.01575102 1.96924907 -0.0230692925557841 GR5J 1.87057608 1.84606507 -0.0131034552735219 GR6J 2.87911398 2.87772704 -0.000481724589451707 CemaNeigeGR4J 4.95347012 21.27312496 3.29459034669619 CemaNeigeGR5J 4.95872498 20.51454101 3.13705964592535 CemaNeigeGR6J 5.97442893 22.58259913 2.7798757663019 CemaNeigeGR4J_Hyst 5.12278008 20.95279899 3.09012268002729 CemaNeigeGR5J_Hyst 5.10475693 21.10790191 3.13494750082057 CemaNeigeGR6J_Hyst 6.06720607 22.35665413 2.68483514027075 GR4H 6.80006994 6.54337394 -0.0377490235048965 GR5H 7.13034895 6.59144698 -0.0755786250825775 CemaNeigeGR4H 21.29121298 100.64233398 3.72694224018795 CemaNeigeGR5H 21.49464107 96.588994 3.49363139795849 CemaNeigeGR4H_Hyst 21.10567399 96.11279095 3.55388399325882 CemaNeigeGR5H_Hyst 21.75877307 103.31538195 3.74821726471547 Execution time is dramatically increased for CemaNeige models only. GRxJ models are all slightly more fast than before which is comforting.
Edited by Dorchies David - Dorchies David mentioned in commit 477a3a2e
mentioned in commit 477a3a2e
- Author Developer
Found!
Here's the value of
RunOptions$Outputs_Cal
in v1.6.12:> RunOptions$Outputs_Cal [1] "PliqAndMelt" "Qsim"
And now in branch in v1.6.12.9000 branch dev:
> RunOptions$Outputs_Cal [1] "DatesR" "PotEvap" "Precip" "Interc" "Prod" "Pn" [7] "Ps" "AE" "EI" "ES" "Perc" "PR" [13] "Q9" "Q1" "Rout" "Exch" "AExch1" "AExch2" [19] "AExch" "QR" "QD" "Qsim" "Pliq" "Psol" [25] "SnowPack" "ThermalState" "Gratio" "PotMelt" "Melt" "PliqAndMelt" [31] "Temp" "Gthreshold" "Glocalmax" "WarmUpQsim" "StateEnd" "Param"
Seem that I didn't manage correctly this variable in
CreateRunOptions
with CemaNeige...Edited by Dorchies David - Dorchies David mentioned in commit e696b80f
mentioned in commit e696b80f
- Dorchies David mentioned in commit b0cf954f
mentioned in commit b0cf954f
- Dorchies David mentioned in commit 593017b3
mentioned in commit 593017b3
- Dorchies David mentioned in commit 82364d6f
mentioned in commit 82364d6f
- Dorchies David mentioned in commit 303ff41a
mentioned in commit 303ff41a
- Delaigue Olivier mentioned in commit b3a73e00
mentioned in commit b3a73e00
- Delaigue Olivier closed via commit e696b80f
closed via commit e696b80f
- Delaigue Olivier closed via merge request !57 (merged)
closed via merge request !57 (merged)
- Dorchies David mentioned in commit 1cabf829
mentioned in commit 1cabf829
- Dorchies David mentioned in commit 9e5a0008
mentioned in commit 9e5a0008
- Dorchies David mentioned in commit 4ff3fce8
mentioned in commit 4ff3fce8
- Dorchies David mentioned in commit c639cd6d
mentioned in commit c639cd6d
- Delaigue Olivier mentioned in commit 3398e7d4
mentioned in commit 3398e7d4