Skip to contents

Extract/plot running time of functions used for different integration methods

Source: R/targets_running_time.R
get_targets_running_time.Rd

In the plot, the coloured blocks in each bar each represent one of the steps (targets) that correspond to an integration method. The text above each bar shows the function that took the longest to run for each integration method.

Usage

get_targets_running_time(
  target_patterns = c(sPLS = "^spls_", sO2PLS = "^so2pls_", MOFA = "^mofa_", DIABLO =
    "^diablo_"),
  patterns_to_methods = c("sPLS", "sO2PLS", "MOFA", "DIABLO"),
  target_exclude_patterns = NULL,
  method_functions = get_method_functions()
)

plot_running_time(
  target_patterns = c(sPLS = "^spls_", sO2PLS = "^so2pls_", MOFA = "^mofa_", DIABLO =
    "^diablo_"),
  patterns_to_methods = c("sPLS", "sO2PLS", "MOFA", "DIABLO"),
  target_exclude_patterns = NULL,
  method_functions = get_method_functions()
)

Arguments

target_patterns

Named character vector, regex patterns used to extract targets by their names.

patterns_to_methods

Character vector of same length as target_patterns, gives the integration method to which each pattern from target_patterns corresponds.

target_exclude_patterns

Character vector, regex used to exclude some targets from the analysis.

method_functions

Named list of character vector, giving for each integration method (i.e. values in patterns_to_methods) the name of the functions used in the pipeline to be considered when computing the running time. Names should match values in patterns_to_methods.

Value

Either a tibble of running time for each function used as part of each integration method (get_targets_running_time), or a ggplot (plot_running_time).

Examples

if (FALSE) { # \dontrun{
## By default uses all targets whose name starts with "spls_" as part of the
## "sPLS" method, etc
get_targets_running_time()
plot_running_time()

## If we ran two versions of MOFA, e.g. one with supervised input and the
## other with unsupervised input
get_targets_running_time(target_exclude_patterns = "^mofa_unsupervised")
plot_running_time(target_exclude_patterns = "^mofa_unsupervised")

## If instead we want to compare the two MOFA run times
get_targets_running_time(
  target_patterns = c(
    "MOFA supervised" = "^mofa_(?!=unsupervised)",
    "MOFA unsupervised" = "^mofa_unsupervised"
  ),
  patterns_to_methods = c("MOFA", "MOFA")
)
plot_running_time(
  target_patterns = c(
    "MOFA supervised" = "^mofa_(?!=unsupervised)",
    "MOFA unsupervised" = "^mofa_unsupervised"
  ),
  patterns_to_methods = c("MOFA", "MOFA")
)
} # }