Attribute-based (such as site) stratification

[raamana]

Given a data structure such pyradigm capable of holding attributes for each sample (such as site, scanner and other meta-data), device an estimating pipeline (CV loop + other infra) that can produce reports on the impact of that attribute.

[raamana]

reasons to do this / evidence from literature:

• Soneson, C., Gerster, S., & Delorenzi, M. (2014). Batch Effect Confounding Leads to Strong Bias in Performance Estimates Obtained by Cross-Validation. PLoS ONE, 9(6), e100335. https://doi.org/10.1371/journal.pone.0100335