Skip to content

processing-guide.md

Latest commit

 

History

History
163 lines (109 loc) · 4.13 KB

processing-guide.md

File metadata and controls

163 lines (109 loc) · 4.13 KB

Processing Guide

中文版本: processing-guide.zh.md

This guide explains how to use the session-based PSRCHIVE processing workflow from the desktop UI.

Core model

  • Opening an archive creates a backend processing session.
  • The renderer reads charts from /api/sessions/{id}/preview/*.
  • Every processing change updates the session recipe, then refreshes the preview.
  • Nothing mutates the source archive in place.
  • Save Archive and batch runs always export new files.

Opening the Processing Inspector

  1. Open any archive file.
  2. In the chart toolbar, click the Processing button.
  3. The right-side Processing Inspector opens with five tabs:
    • Zap
    • Pam
    • TOA
    • Cal
    • Batch

If the backend is running in mock mode, or required PSRCHIVE tools are missing, the panel stays visible but shows capability notes instead of active controls.

Zap workflow

v1 supports channel zapping from the Freq x Phase waterfall.

Single channel

  1. Switch to Freq x Phase.
  2. Click a channel row in the heatmap.
  3. The channel is added to recipe.zap.channels.
  4. Click the same channel again to unzap it.

Channel range

  1. In the Plotly modebar, choose box select.
  2. Drag vertically across one or more channel rows.
  3. The selected channels are merged into recipe.zap.channels.

The Zap tab lists every selected channel and lets you remove them individually or clear the whole set.

Pam controls

The Pam tab exposes live v1 controls:

  • dedisperse
  • tscrunch
  • fscrunch
  • bscrunch
  • phase rotate

Notes:

  • Slider changes are debounced before the backend preview refreshes.
  • Preview uses the same recipe that export uses.
  • dedisperse changes chart extraction immediately and is also applied during archive export.

TOA extraction

The TOA tab is a v1 pat workflow.

  1. Choose a template archive.
  2. Pick an algorithm (PGS, GIS, PIS, SIS, ZPS).
  3. Choose output format (tempo2 or parkes).
  4. Optionally enable time/frequency scrunch before TOA extraction.
  5. Optionally choose an output text path.
  6. Click Run TOA.

The result area shows:

  • parsed TOA rows
  • raw pat output
  • a visual residual chart with:
    • observed profile
    • aligned template
    • difference trace

This is not yet a full tempo2 timing residual workflow.

Calibration preview

The Cal tab is a v1 pac workflow for existing calibration assets.

You can provide:

  • a calibration search path
  • database.txt
  • a solution file

You can also choose:

  • model: SingleAxis, Polar, Reception
  • pol-only

The current session preview updates with calibration enabled, and Refresh shows the latest pac command and log output.

Exporting the current archive

Save Archive exports the current session recipe to a new archive path.

The default filename pattern is:

<original-stem>.<archiveExtension><original-extension>

Example:

J0437-4715.processed.ar

The output never overwrites the source file unless you explicitly choose the same target path in the save dialog.

Batch recipes

The Batch tab does three things:

  1. stores reusable recipes per workspace
  2. controls default export naming/output directory
  3. runs sequential batch exports

Save a recipe

  1. Build the live recipe with the Zap, Pam, TOA, and Cal tabs.
  2. Enter a recipe name in Batch.
  3. Click Save recipe.

Reuse a saved recipe

  1. In Batch, click Load on a saved recipe.
  2. The current session preview updates to match that recipe.

Run a batch export

  1. In Batch, choose whether TOA text should be exported alongside archives.
  2. Optionally set an output directory.
  3. Click Select files and run.
  4. Pick one or more archive files.

The app then, for each file:

  1. creates a temporary processing session
  2. applies the selected recipe
  3. exports a processed archive copy
  4. optionally exports a TOA text file
  5. records success or failure in the batch log

Recommended runtime

Use Docker / OrbStack for the most complete workflow:

npm run backend:docker:pull
npm run dev:docker

That runtime gives the app stable access to paz, pam, pat, pac, and tempo2.