Articles
Vignettes
- Fitting a proxy to a density you cannot sample
- Compressing a Bayesian posterior you can evaluate but not sample
- Choosing between the three fitting regimes
- How well a mixture proxies four awkward shapes
- The closed-form operator calculus on a mixture
- Compressing a kernel density estimate into a mixture
- One mixture, many methods
- Reading the entropy of a fitted mixture
- Mapping the optima of an objective
- Imputing missing data with a mixture
- Missing data that depends on the missing value
- Testing the last observation for instability
Extended articles
Longer versions of the vignettes. Each compares proxymix with other R packages on the same task, using stored simulation results.
- Fitting a proxy to a density you cannot sample
- Compressing a Bayesian posterior you can evaluate but not sample
- Choosing between the three fitting regimes
- How well a mixture proxies four awkward shapes
- The closed-form operator calculus on a mixture
- Compressing a kernel density estimate into a mixture
- One mixture, many methods
- Reading the entropy of a fitted mixture
- Mapping the optima of an objective
- Imputing missing data with a mixture
- Missing data that depends on the missing value
- Testing the last observation for instability