delete old replace with new script

[nhsavage]

replace the two old scripts

also move all user configurable options to top of script

[github-actions]

Coverage Report

File	Stmts	Miss	Cover	Missing
lib/osop
compare_terciles.py	77	77	0%	9–331
compute_products_func.py	135	135	0%	8–449
compute_scores_func.py	155	155	0%	17–415
ens_plotting.py	117	38	68%	106–111, 197, 270–289, 309–356
plot_verify.py	158	131	17%	88–110, 142–228, 255–288, 318–410, 434–511
pycpt_convert.py	394	287	27%	76–99, 116–122, 140, 163–169, 189–205, 219–225, 242–270, 290, 311–313, 340–531, 554, 569–571, 663, 736–744, 759–765, 776–791, 796–799, 804–808, 816–827, 843–897, 917–971, 1009–1044
run_pycpt.py	126	126	0%	11–362
util.py	59	12	80%	89, 103, 131, 159–172, 179–184
TOTAL	1537	961	37%

Tests	Skipped	Failures	Errors	Time
31	0 💤	0 ❌	0 🔥	12.950s ⏱️

[Copilot]

Pull request overview

This PR consolidates workflow execution by replacing the standalone forecast script with an updated run_all.sh, and moves user-configurable options to the top of the script for easier customization.

Changes:

• Adds a user-configuration block at the top of run_all.sh, including toggles to run hindcast and/or forecast.
• Extends run_all.sh to create forecast directory structure and execute forecast download/products/plot steps.
• Deletes scripts/master.forecast.sh as part of replacing old scripts.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.

File	Description
scripts/run_all.sh	Centralizes configuration, adds do_hc/do_fc flow control, and integrates forecast execution into the main run script.
scripts/master.forecast.sh	Removed in favor of the consolidated run_all.sh workflow.

Comments suppressed due to low confidence (3)

scripts/run_all.sh:31

• base_path is computed from $SCRATCH before any validation, and set -u is enabled later. If SCRATCH is unset/empty, this silently writes into /osop/... (or another unintended location), which is risky on shared systems. Consider asserting SCRATCH is set and quoting path construction.
  scripts/run_all.sh:72
• Unquoted path variables in mkdir -p can word-split or glob if any directory component contains spaces or wildcard characters, and can also mis-handle paths beginning with -. Quoting and using -- makes these directory creations more robust.
  scripts/run_all.sh:125
• cp should quote paths (and ideally use an explicit destination filename) to avoid word-splitting/globbing issues and to make it clear the intent is to copy the file into the forecast downloads directory.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated no new comments.

Comments suppressed due to low confidence (3)

scripts/run_all.sh:86

• PYTHONPATH export will prepend a leading ":" when PYTHONPATH is unset/empty, which implicitly adds the current working directory to sys.path. That’s a security footgun and also regresses the safer pattern previously used in this script.
  scripts/run_all.sh:261
• In the forecast path, the script continues into forecast_products/forecast_plots even when the per-centre download fails. This will typically cascade into avoidable failures/noisy logs; the hindcast path already continues on download failure, so forecast should mirror that behavior.
  scripts/run_all.sh:126
• With do_hc=0, the script still rewrites $downloaddir/parseyml.yml and then copies it into the forecast downloads. That can break “forecast-only” runs that rely on previously-downloaded hindcast data, because get_any_hindcast.py / forecast_products.py use parseyml.yml to determine system IDs and will now see the new mapping rather than the one that matches existing hindcast files.

[Copilot]

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated no new comments.

Comments suppressed due to low confidence (3)

scripts/run_all.sh:262

• In the forecast branch, a failed download currently falls through to forecast_products.py / forecast_plots.py. That can waste runtime and can produce misleading downstream errors when inputs are missing. Consider skipping the remainder of the loop for this centre when the download step fails (mirrors the hindcast path above).
  scripts/run_all.sh:287
• This error message still references running master.sh, but this script now combines hindcast+forecast and is intended to replace the older scripts. Updating the message to describe the actual prerequisite (hindcast products/terciles exist for the same settings) will make failures easier to diagnose.
  scripts/run_all.sh:16
• This PR deletes scripts/master.forecast.sh, but there are still user-facing references to that script elsewhere (e.g. docs and error messages). This will mislead users and can complicate troubleshooting. Please update those references to point to the new workflow/script name.

[nhsavage]

also makes the default no borders

[nhsavage]

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated no new comments.

Comments suppressed due to low confidence (3)
scripts/run_all.sh:262

• In the forecast branch, a failed download currently falls through to forecast_products.py / forecast_plots.py. That can waste runtime and can produce misleading downstream errors when inputs are missing. Consider skipping the remainder of the loop for this centre when the download step fails (mirrors the hindcast path above).
  scripts/run_all.sh:287
• This error message still references running master.sh, but this script now combines hindcast+forecast and is intended to replace the older scripts. Updating the message to describe the actual prerequisite (hindcast products/terciles exist for the same settings) will make failures easier to diagnose.
  scripts/run_all.sh:16
• This PR deletes scripts/master.forecast.sh, but there are still user-facing references to that script elsewhere (e.g. docs and error messages). This will mislead users and can complicate troubleshooting. Please update those references to point to the new workflow/script name.

opened #165 to check these