search

JPL-MGHG / JPL-MGHG-SDS

This is a generic repo for tracking of SDS tasks
0 stars 0 forks source link

Run a Single Version of DALEC for Benchmarking #30

Open ddalton-swe opened 5 months ago

aantonyb commented 5 months ago

(1) Here's the command: CARDAMOM_RUN_MODEL.exe PROJSCRIPT_CARDAMOM_CMSFLUX_PRIORS_JUL23_V1_1942.nc PROJSCRIPT_CARDAMOM_CMSFLUX_PRIORS_JUL23_V1_1942_SINGLE_PAR.cbr local_output.nc (2) The first two files (input files) are available on the CARDAMOM teams folder. (3) you can then compare "local_output.nc" with "PROJSCRIPT_CARDAMOM_CMSFLUX_PRIORS_JUL23_V1_1942_SINGLE_PAR.nc", the latter is also on the CARDAMOM teams folder

ddalton-swe commented 5 months ago

@aantonyb I got the same value for both

Median: -23.53569009067604
25th Percentile: -23.53569009067604
75th Percentile: -23.53569009067604
aantonyb commented 5 months ago

OK great! For the single run, can you show the differences between all the FLUXES values (between your run and the control run) and same for all the POOLS values? How big are the differences between your run and the control run?

E.g. below is the full set of FLUXES values. If you take the difference, i.e. “FLUXES_your_new_run – FLUXES_teams_folder_control_run”, are the values zero? If yes, then success! I expect them to be ~1e-10 or something like that

@.***

From: Dillon Dalton @.> Date: Tuesday, January 30, 2024 at 09:29 To: JPL-MGHG/JPL-MGHG-SDS @.> Cc: Bloom, Anthony (US 329F) @.>, Mention @.> Subject: [EXTERNAL] Re: [JPL-MGHG/JPL-MGHG-SDS] Run a Single Version of DALEC for Benchmarking (Issue #30)

@aantonybhttps://urldefense.us/v3/__https:/github.com/aantonyb__;!!PvBDto6Hs4WbVuu7!IIiMgZTif9B_qOL2zLIgZPOpaPyLNDK4w8nXcEK5b897ZnvgdnP6g4CJVRR9pmbcB2NQTTrc259xkXqlH4zMRdtEFLw$ I got the same value for both

Median: -23.53569009067604

25th Percentile: -23.53569009067604

75th Percentile: -23.53569009067604

— Reply to this email directly, view it on GitHubhttps://urldefense.us/v3/__https:/github.com/JPL-MGHG/JPL-MGHG-SDS/issues/30*issuecomment-1917545301__;Iw!!PvBDto6Hs4WbVuu7!IIiMgZTif9B_qOL2zLIgZPOpaPyLNDK4w8nXcEK5b897ZnvgdnP6g4CJVRR9pmbcB2NQTTrc259xkXqlH4zMfusYdlY$, or unsubscribehttps://urldefense.us/v3/__https:/github.com/notifications/unsubscribe-auth/AFTYZNDC2XI55VDFVI7HWIDYREUW7AVCNFSM6AAAAABCQBDQ32VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMJXGU2DKMZQGE__;!!PvBDto6Hs4WbVuu7!IIiMgZTif9B_qOL2zLIgZPOpaPyLNDK4w8nXcEK5b897ZnvgdnP6g4CJVRR9pmbcB2NQTTrc259xkXqlH4zMxEjMyjM$. You are receiving this because you were mentioned.Message ID: @.***>

ddalton-swe commented 5 months ago

These are the reported differences

Mean Absolute Difference of POOLS: 6.609699975792169e-14
Mean Absolute Difference of FLUXES: 5.382254113572629e-17
aantonyb commented 5 months ago

That meets requirements and expectations! Do you want to share the output nc file on the teams folder? That way, if we want to look into more diagnostics, we have the files at hand.

Thanks!

From: Dillon Dalton @.> Date: Tuesday, January 30, 2024 at 10:27 To: JPL-MGHG/JPL-MGHG-SDS @.> Cc: Bloom, Anthony (US 329F) @.>, Mention @.> Subject: [EXTERNAL] Re: [JPL-MGHG/JPL-MGHG-SDS] Run a Single Version of DALEC for Benchmarking (Issue #30)

These are the reported differences

Mean Absolute Difference of POOLS: 6.609699975792169e-14

Mean Absolute Difference of FLUXES: 5.382254113572629e-17

— Reply to this email directly, view it on GitHubhttps://urldefense.us/v3/__https:/github.com/JPL-MGHG/JPL-MGHG-SDS/issues/30*issuecomment-1917643495__;Iw!!PvBDto6Hs4WbVuu7!Lso6oGfvPu1tmxYL0amatpej43Zv4U7YxDYSweU5I3oxoqVtYySBBiLV0Wllyfw0EhsfrCnTBO-eUWv73q906mYrYVo$, or unsubscribehttps://urldefense.us/v3/__https:/github.com/notifications/unsubscribe-auth/AFTYZNE6XJSMRCV5CTGGYP3YRE3PTAVCNFSM6AAAAABCQBDQ32VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMJXGY2DGNBZGU__;!!PvBDto6Hs4WbVuu7!Lso6oGfvPu1tmxYL0amatpej43Zv4U7YxDYSweU5I3oxoqVtYySBBiLV0Wllyfw0EhsfrCnTBO-eUWv73q90QbNRYCg$. You are receiving this because you were mentioned.Message ID: @.***>

ddalton-swe commented 5 months ago

Good idea. I have added it