R/MethodsHandling.R

In IsamBenS/3CClusteringTool: Clustering Tool

library(microbenchmark)

run.algo.param.combi <- function(p.list)
{
    run.algo.param.combi.output.values.list <- list()
    
    recursive.combi.function <- function(params.list, values)
    {
        val.list <- unlist(values)
        if(length(params.list)>0)
        {
            lapply(params.list[[1]], function(val)
            {
                tmp <- c(val.list, val)
                if(length(params.list)>1)
                {
                    recursive.combi.function(params.list[2:length(params.list)], tmp)
                }
                else
                {
                    recursive.combi.function(NULL, tmp)
                }
            })
        }
        else
        {
            run.algo.param.combi.output.values.list[[length(run.algo.param.combi.output.values.list)+1]] <<- val.list
        }
    }
    
    recursive.combi.function(p.list, NULL)
    return(run.algo.param.combi.output.values.list)
}

run.algo.combi <- function(p.list)
{
    run.algo.param.combi.output.values.list <- list()
    
    recursive.combi.function <- function(params.list, values)
    {
        val.list <- values
        if(length(params.list)>0)
        {
            lapply(params.list[[1]], function(val)
            {
                tmp <- c(val.list, val)
                if(length(params.list)>1)
                {
                    recursive.combi.function(params.list[2:length(params.list)], tmp)
                }
                else
                {
                    recursive.combi.function(NULL, tmp)
                }
            })
        }
        else
        {
            run.algo.param.combi.output.values.list[[length(run.algo.param.combi.output.values.list)+1]] <<- val.list
        }
    }
    
    recursive.combi.function(p.list, NULL)
    return(run.algo.param.combi.output.values.list)
}

benchmark.method <- function(method.name, fcs.file, parameters = list(), markers_cols)
{
    fct.name <- paste0("BRP_BM.",method.name,".execute")
    fct <- match.fun(fct.name, descend = FALSE)
    fcs.output <- NULL
    time.output <- mean(microbenchmark({    fcs.output <- fct(fcs.file, parameters, markers_cols)    }, times = 1, unit = "ns")$time)

    method.output <- list(fcs.output, time.output)

    return(method.output)
}

benchmark.source.method <- function(method.dir, method.name)
{
    source(paste0(method.dir,"/",method.name,".R"))
}

# save.exec.time <- function(method.name, directory="", time.output)
# {
#     times.vector <- NULL
#     if("times.csv" %in% list.files(directory))
#     {
#         times.vector <- c(as.matrix(read.csv(paste0(directory,"times.csv")))[1,])
#         if(!(method.name %in% names(times.vector)))
#         {
#             times.vector[[length(times.vector) + 1]] <- time.output
#             names(times.vector)[length(times.vector)] <- method.name
#         }
#         else
#         {
#             times.vector[[method.name]] <- time.output
#         }
# 
#     }
#     else
#     {
#         times.vector <- time.output
#         names(times.vector) <- method.name
#     }
#     mat <- matrix(times.vector, nrow = 1)
#     colnames(mat) <- names(times.vector)
#     if(file.exists(paste0(directory,"times.csv")))
#     {
#         file.remove(paste0(directory,"times.csv"))
#     }
#     write.csv(mat, paste0(directory,"times.csv"), row.names = FALSE)
# }
# 
# read.exec.time <- function(method.name, directory="")
# {
#     time.output <- NULL
#     if("times.csv" %in% list.files(directory))
#     {
#         temp <- as.vector(read.csv(paste0(directory,"times.csv")))
#         if(method.name %in% names(temp))
#         {
#             time.output <- temp[method.name]
#         }
#     }
#     
#     return(time.output)
# }