plotParamEvolutionMsg: Plot the value of a parameter along the optimisation process

In jmp75/mhplot: Visualise outputs from an optimisation algorithm

Plot the value of a parameter along the optimisation process

Description

Plot the value of a parameter along the optimisation process. The color scale is the message associated with the point, for instance a geometric transformation. An example message is 'Contracted point in the subcomplex' for the Shuffled complex algorithm. Useful to check the behavior of the optimisation process.

Usage

plotParamEvolutionMsg(logMh, paramName)

Arguments

logMh	an mhData object
paramName	the exact name of one of the model parameters

Value

a ggplot object

Examples

## Not run: 
logFileName <- 'F:/path/to/419016_IhacresClassic_19390101_19481231.csv'
logSce <- loadMhLog(logFileName)
logMh <- new("mhData",
   data= logSce,  
   fitness = "NSE.logbias",
   messages = "Message", 
   categories = "Category")
   
geomOps <- subsetByMessage(logMh)
d <- plotParamEvolutionMsg(geomOps, 'Tq')
d
# If there is overplotting, one can use facets to have a clearer view of the optimiser behavior
d + facet_wrap( as.formula(paste("~", geomOps@messages, sep=' ')) )

## End(Not run)

jmp75/mhplot documentation built on Feb. 4, 2023, 9:31 p.m.