Skip to content

output-report.md

Latest commit

 

History

History
90 lines (70 loc) · 3.99 KB

output-report.md

File metadata and controls

90 lines (70 loc) · 3.99 KB
Raw

Output report

After a pipeline run (or against any existing MEA-NAP output folder), the 🌐 View report button on the Results tab generates report.html at the root of that output folder and opens it in your browser.

If the run used {doc}`express-mode`, that button opens the run's `.meanap`
bundle in the interactive viewer instead — an express run keeps only the bundle,
so there is no folder to build a report from. The viewer's **Export output
folder** writes one, `report.html` included. See [Related: the interactive
viewer](#related-the-interactive-viewer) below.

# You can also generate it directly, without the GUI:
from meanap.pipeline.report import generate_report
generate_report("/path/to/OutputData...")  # writes report.html there, returns its path

What it looks like

  • A folder-tree sidebar on the left, matching the same output structure MATLAB's CreateOutputFolders.m builds (1_SpikeDetection, 2_NeuronalActivity, 3_EdgeThresholdingCheck, 4_NetworkActivity, ...).
  • A captioned image gallery on the right for whichever folder is selected.
  • Data files (.npz/.json/.csv/.mat) are listed with a short caption rather than embedded — clicking one opens/downloads it via your browser's normal file:// handling.

It is a single self-contained HTML file: no server, no external JavaScript or CSS, no new dependencies beyond a browser. It works entirely offline and can be emailed, zipped, or committed alongside the rest of an output folder.

Where captions come from

Figure captions are adapted from MEA-NAP's own figure-legend reference (MATLAB outputs) wherever that page documents a matching figure, reworded to describe what the Python port's version of the plot actually shows (MATLAB's originals sometimes also render additional "scaled to whole dataset" or "combined" variants the Python port doesn't produce). A handful of step-2 burst-heatmap figures have no MATLAB documentation anywhere in the repository; their captions were written from scratch to match the documented semantics of their sibling figures.

Deep links

Every plot lives at a URL fragment you can share directly:

report.html#4_NetworkActivity/4A_IndividualNetworkAnalysis/<group>/<recording>/<lag>mslag

Opening a link like this auto-expands the sidebar tree and navigates straight to that folder — useful for pointing a labmate at one specific plot without walking them through the tree by hand.

Run parameters

The sidebar's ⚙ Run parameters entry shows the settings the run used, read from params.json and grouped the way Params groups them — Recording, Spike detection, Connectivity, and so on. It opens on only what differs from the defaults, which on a typical run is a dozen fields out of ~140, with the default each one departed from shown beside it; a toggle expands to all of them.

Remote share links are replaced with a placeholder. A report is a file people attach to papers and email onward, and a Dropbox link in one is a credential.

The entry is absent for an output folder with no params.json — an older run, or one that failed before writing it.

Related: the interactive viewer

report.html shows the figures a run wrote. If you ran with {doc}express-mode, none were — use meanap-viewer instead, which redraws them on demand and can export SVG. The two are complementary: report.html needs nothing installed, the viewer needs Python running but can restyle and re-export.

To get a report.html out of an express run, open the bundle and press Export output folder: it draws every figure into a normal output folder and generates the report alongside them. That folder is what you send to someone without MEA-NAP.

The GUI picks between them for you: 🌐 View report opens the viewer for an express run and builds report.html for a full one. You can also open any bundle directly with 📦 Open bundle… on the Results tab, or by dragging the .meanap file onto the window.