FLASH-71 Refactor flashfinder for new modules

[FatemehMoghadam]

Replaced the old module set (which required manually swapping to pawseyenv/2024.05 and loading gcc/12.2.0 and libfabric) with the new module set that works natively in the default pawseyenv/2025.08 environment.

scipy.integrate.simps was deprecated in SciPy 1.11 and completely removed in SciPy 1.14. Replaced both calls with scipy.integrate.simpson, and updated the x axis argument to be passed as a keyword argument (x=rest_vel)

Activated the existing __float_from_str() helper function (which was already in the codebase but commented out) to handle Fortran scientific notation.

On second thought we should definitely run a comparison of outputs between the updated modules pipeline and the existing pipeline before merging this

[ree13d]

Can we before merging this @FatemehMoghadam somehow run a comparison of outputs for a small set of SBIDs between the updated modules pipeline and the existing pipeline? We have no unit tests so if say for example the outputs of the various packages used are slightly different we would impact science outputs.

So the MR is fine in general, but im worried we dont know if the upgrades change the results.

[gwhgerman]

OFFICIAL Just to note: the FLASH backend is based on statistical algorithms, meaning that no two runs will produce exactly the same results. The test of the backend outputs will need the science team to review the outputs and give a considered opinion on whether they're acceptable. I'm putting together a document on the process and will upload it to Confluence once I've chatted to Hyein and Liz. Cheers, Gordon

…

---------------------------------------- Dr Gordon German Head of Software Development, Australian Regional SKA Centre, Perth, Western Australia

________________________________ From: ree13d ***@***.***> Sent: Thursday, 30 April 2026 12:18 PM To: AusSRC/FLASH ***@***.***> Cc: German, Gordon (S&A, Kensington WA) ***@***.***>; Review requested ***@***.***> Subject: Re: [AusSRC/FLASH] FLASH-71 Refactor flashfinder for new modules (PR #14) [https://avatars.githubusercontent.com/u/259253567?s=20&v=4]ree13d left a comment (AusSRC/FLASH#14)<#14 (comment)> Can we before merging this @FatemehMoghadam<https://github.com/FatemehMoghadam> somehow run a comparison of outputs for a small set of SBIDs between the updated modules pipeline and the existing pipeline? We have no unit tests so if say for example the outputs of the various packages used are slightly different we would impact science outputs. — Reply to this email directly, view it on GitHub<#14 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ARIULNVMHFDRLFTETOQTIHL4YLHZJAVCNFSM6AAAAACX27SV5KVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DGNBZGY2TCMRUHE>. Triage notifications on the go with GitHub Mobile for iOS<https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android<https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because your review was requested.Message ID: ***@***.***>