Reproducibility¶

CloudScope is designed so desktop GUI workflows, browser GUI workflows, scripts, and notebooks use the same acqstore scientific backend.

This is important for scientific reproducibility because analysis behavior should not depend on whether a user clicks through the GUI or calls the API from Python.

Versioned releases¶

Official CloudScope releases are tied to git tags and published on the CloudScope Releases page.

A published release provides a reproducible reference point for scientific analysis while allowing active development to continue.

Release artifacts can include:

source code archives
desktop application builds
documentation builds
version metadata

Sidecar files¶

CloudScope saves analysis state and results next to the source image file.

For a source file named my_file.tif, saved files may include:

my_file.tif.json
my_file.tif.radon_velocity.csv
my_file.tif.diameter.csv

The JSON sidecar records the ROIs, detection parameters, and analysis summaries used to generate results. These files should be retained with the source data when preserving an analysis.

Same backend across interfaces¶

CloudScope uses the same acqstore backend from:

desktop GUI
browser GUI
Python scripts
Jupyter notebooks

This reduces divergence between interactive and scripted workflows.