get_wait_cor_summary: Generate summary dataframe of correlations with...
In robjohnnoble/demonanalysis: Analysis and plotting of demon data

Description Usage Arguments Value Examples

Generate summary dataframe of correlations with "waiting_time"

1
2
3

get_wait_cor_summary(summary, col_names_list, num_parameters, min_count,
  Verbose = FALSE, ReturnCI = FALSE, VariablesToNotGroupBy = NULL,
  method = "spearman")

`summary`	dataframe including columns named "seed", "Generation", "start_time", "start_size", "gap" and "waiting_time"
`col_names_list`	char vector of column names in the summary dataframe
`num_parameters`	number of parameters, accounting for the first set of columns in the dataframe
`min_count`	minimum number of items in each column (otherwise result will be NA)
`Verbose`	if TRUE, helpful to debug, print the name of the variables with which compute the correlation
`ReturnCI`	if true, also return the 0.95 level confidence interval computed by bootstraping.
`VariablesToNotGroupBy`	vector of column names by which we don't want to group the simulation. For example, if mutation rate (mu_driver_birth) are random, then we don't want to group the simulations by the variable mu_driver_birth, so we need to set VariablesToNotGroupBy=c("mu_driver_birth") This is more general than creating a boolean argument as CombinedMutationRate=TRUE/FALSE.

Dataframe with one row for each unique combination of parameter values and start_size (i.e. it summarises over "seed"), and including columns containing the correlations between "waiting_time" and each variable in col_names_list and the associated pValues for the two.sided test of the correlation coefficient. If the argument ReturnCI=TRUE, the 0.95 Confidence Intervals for the correlation coefficients are also computed. Argument VariablesToNotGroupBy allows to compute the correlation coefficients while not grouping simulations by variables contained into VariablesToNotGroupBy.

wait_cor_summary <- get_wait_cor_summary(sum_df, 
c("DriverDiversity", "DriverEdgeDiversity"), 15, min_count = 2)
depth_wait_cor_summary <- get_wait_cor_summary(sum_df, 
c(paste0("DriverDiversityFrom1SamplesAtDepth", 0:10), 
paste0("DriverDiversityFrom4SamplesAtDepth", 0:10)), 
15, min_count = 2)