Functions and utilities that extend the functionality of the opal/datashield suite of packages. Server side.
dsSwissKnife is a software package which builds upon (and extends) DataSHIELD (https://www.datashield.ac.uk) to allow non-disclosive remote federated analysis on sensitive data. Like most DataSHIELD packages it has a client-server infrastructure so it needs to be used together with the client version, dsSwissKnifeClient (https://github.com/sib-swiss/dsSwissKnifeClient).
Functionality include implementations for a kmeans clustering algorithm, PCA, as well as a number of functions from specific packages (imputation with VIM, random forests with randomForest, synthetic data generation with Synthpop, etc). Please consult the documentation for the individual functions in dsSwissKnifeClient for more details.
Some utilisation examples and a basic docker setup are available here: https://github.com/sib-swiss/dsSwissKnife-example