Skip to content

Comparing RS — Python tools UI (web-based)

A browser-based interface for the Surrogate MDS comparison — no installation required to use the hosted version below.

Live app

App URL
Surrogate MDS Analysis eih27-rs-software-app-streamlit-ui-n69iyq.streamlit.app

Note: the app may be asleep if inactive — click the link and wait a moment for it to wake up. The hosted version runs on limited, shared resources — it's meant for small datasets and a quick look, not a final fit.

Running locally

Faster than the hosted version, and not limited by its shared resources:

cd rs-software
pip install streamlit plotly pandas matplotlib scikit-learn
streamlit run app.py

Then open http://localhost:8501 in your browser.

How to use

Compare two choice datasets and test whether the difference in their perceptual maps is statistically meaningful.

  1. Upload two .mat choice files (or enter their file paths)
  2. Set dimensions, number of surrogates, and resampling method
  3. Click Run analysis

Output:

  • Log-likelihood and Procrustes disparity for each dataset
  • Convergence plot (log-likelihood per iteration)
  • Null distribution of surrogate disparities vs. real disparity
  • p-value: fraction of surrogates with disparity ≥ real

Resampling methods:

Method Description
With replacement Draw observations with replacement from pooled data (default)
Without replacement Draw without replacement from pooled data
Without replacement, paired Draw without replacement, keeping A and B draws non-overlapping

Random seed: the sidebar has a "Random seed" control for the real-data fit specifically (surrogates are already seeded deterministically per-surrogate) — choose "Random each run" for normal use, or "Same every run" / "Custom random seed" for a reproducible result.

For the equivalent Python function, see Comparing RS → Python.