spaCR main · Main documentation · Nightly preview
Contents Menu Expand Light mode Dark mode Auto light/dark, in light mode Auto light/dark, in dark mode Skip to content
spaCR 1.5.1.6 documentation
Logo
spaCR 1.5.1.6 documentation
  • Installer guide
  • System requirements
  • Navigator: choose a tool from your input
  • Choose a workflow after installation
  • Combine measurement tables for plots and gates
  • Installer archive
  • Capabilities
  • Keyboard shortcuts and settings templates
  • Make Masks: editing, detection and measurement
  • Measure: preview checked images and verify mask planes
  • Train a Cellpose model
  • Process images with a point-spread function
  • Recruitment: compartment ratios and channel identity
  • Image quality before segmentation
  • Host–Pathogen Analysis
  • Plaque Assay: fields, figures and reviewed conditions
  • Python API quickstart
  • Export measurements to AnnData
  • Where a setting goes
  • Model zoo
  • Language
  • Setting animation gallery
  • Checkpoint and resume
  • Reproducibility manifests
  • Unified run history
  • Plate and batch-effect correction
  • Classifier evaluation workbench
  • Plate-aware guide permutation analysis
  • Explain CV models and investigate hits
  • Resumable multi-objective UMAP search
  • Remote and distributed execution
  • spaCR plugin SDK
  • Train/test leakage audit
  • Threading and cancellation audit
  • Database concurrency audit
  • API reference
    • spacr
      • spacr.api
      • spacr.core
      • spacr.measure
      • spacr.deep_spacr
      • spacr.sequencing
      • spacr.ml
      • spacr.artifacts
      • spacr.settings
Back to top
View this page
Edit this page

Navigator: choose a tool from your input¶

Navigator is the default Home for a fresh profile; a saved opt-out is respected. Launch it with spacr-nav or spacr-navigator, or enable Navigator in Help. Explicit module launches still open that module.

Find a starting point¶

Drop a local file or folder, browse for one, or type a path or task and press Enter. Typing alone does not inspect data or open a module. For example, “detect features” offers Measure; tool names and supported task phrases match without case sensitivity. Recognized commands, such as opening tutorials, can follow their own command route on Enter.

Suggestions appear as a grid of named cards. The first click anchors the selected card and reveals a sparse workflow; it does not run analysis. Hover to read the fixed, scrollable description. A second click on the selected tool opens its native screen. Review its settings and input before pressing Run.

Explore without running¶

Select a downstream card to reveal its branches. More next steps exposes additional direct choices; All tools includes the other available registered tools and subtools. Availability follows maturity and species preferences. A connector describes a workflow relationship, not proof that the current input satisfies every prerequisite. Modes such as 3D, Time and OPS label connections rather than inventing separate processing steps.

Use Back or Escape to return to suggestions. Returning Home starts a fresh Navigator visit. The original card stays anchored while branches grow; the description remains below, and the app dock and side panes can be collapsed without changing the input.

Check what the input supports¶

Local discovery uses bounded filenames, headers and project metadata, not a full analysis. An explicit raw image stays scoped to that input; a verified generated array inside a canonical merged folder can inherit its plate context. Handoffs use the selected file or project root required by the destination. An annotation-only database is not evidence of measured features, and a format keyword is not a promise that every file with that suffix can be read. Missing readers or calibrated-data prerequisites still need attention in the native workflow.

An offered filename-pattern editor requires explicit validation before its pattern is passed onward. Navigator does not silently convert data, run models or complete scientific prerequisite review. Optional AI assistance sends explicitly submitted unmatched questions only when enabled with a chosen provider; local tool and path matches retain their normal routes.

Desktop entry points and API reference¶

The launchers use spacr.qt.run(). Registered screen metadata is described by spacr.qt.app_catalog.DeclaredApp and retrieved with spacr.qt.app_catalog.declared_app(); spacr.qt.app_catalog.register_declared() registers an existing declaration. Resolving spacr.qt.app.registered_factory() can import its screen. These are desktop registry contracts, not a headless Navigator inference API.

For scripted processing, use the supported entry points in Python API quickstart. For the files and prerequisites of each scientific step, see Choose a workflow after installation.

Next
Choose a workflow after installation
Previous
System requirements
Copyright © 2025-2026, Einar Birnir Olafsson
Made with Sphinx and @pradyunsg's Furo
On this page
  • Navigator: choose a tool from your input
    • Find a starting point
    • Explore without running
    • Check what the input supports
    • Desktop entry points and API reference