docs/MASTERPLAN.md

In lenarddome/psp: Parameter Space Partitioning MCMC for Global Model Evaluation

Planned Features

Currently, there are no additional features of parameter space partitioning that can be implemented. Nonetheless, all suggestions are welcome. Keep in mind that there are various features we will not be implementing - calculating volume and burnout.

Some aspects of psp we would like to improve are:

• enhancing performance
• adding various statistics for summarizing psp results
• finding methods to visualise psp outputs

TODO

• :sparkles: init can be a matrix. Users should be able to define multiple jumping distribution.
• :zap: model evaluations should be able to happen in parallel
• :sparkles: connect to external databases
• :zap: allow sampling procedure where parameters are not uniformly distributed

lenarddome/psp documentation built on Dec. 1, 2023, 4:01 a.m.