asd-def#

Investigations#

1. Reproducing the reference classes

Do the four classes reproduce on the release held here, and can they be named as the authors named them?

Reproducing the reference classes
2. Selecting the number of classes

How many classes do the data support, and do the selection criteria back the authors’ choice of four?

Selecting the number of classes
3. Stability under refitting

Do the classes survive re-initialisation and resampling, or are they an artefact of one fit or one sample?

Stability under refitting
4. The minimum viable stratum size

How small can a stratum be before four-class recovery breaks down, which bounds the stratification bins?

The minimum viable stratum size
5. Replicating in the SSC

Do the four classes reappear in a second, independent cohort?

Replicating in the SSC

Packages#

analysis

Reproduce the Litman autism classes and test their stability across age at diagnosis and diagnostic era.

analysis
figures

Render the analysis artefacts as Matplotlib figures for the manuscript.

figures
dscat

A searchable catalogue over the versioned SPARK and SSC data dictionaries.

dscat
docs

Build and live-preview this Sphinx site.

docs