plot_stats {poolABC}

Plot the fit of a summary statistic to the target

Description

Plot the fit of a summary statistic to the target

Usage

plot_stats(sumstat, target, accepted, index = NA, colour = TRUE)

Arguments

sumstat

is a vector or matrix of simulated summary statistics. If this input is a vector, then each entry should correspond to a different simulation. If it is a matrix, then each row should be a different simulation and each column a different statistic. Note that this should be the entire set of simulated values.

target

is an integer or a numeric vector containing the target of the parameter inference. If a single integer, then this should be the target summary statistic corresponding to the input sumstat vector. If this input is a vector, then the order of the entries in the vector should be the same as the order of the columns of the sumstat matrix input. Either way, this input should contain the value of the summary statistics calculated from observed data.

accepted

is a vector or matrix of accepted summary statistics. If this input is a vector, then each entry should correspond to a different simulation. If it is a matrix, then each row should be a different simulation and each column a different statistic. Note that this should be summary statistics of the accepted simulations during parameter inference.

index

is an optional non-negative integer. This input is only required when the sumstat and accepted inputs are matrices. In that instance, it will indicate which summary statistic to plot. It corresponds to the desired column of the sumstat and accepted matrices and to the entry of the target vector.

colour

logical, indicating whether the plot should be a colour version (default) or a grayscale plot.

Value

a plot with the fit of the simulated summary statistics to the observed value. Both the density estimation of the entire simulated summary statistics and the accepted summary statistics are contrasted with the observed value.

Examples

# load the matrix with parameter values
data(params)
# load the matrix with simulated parameter values
data(sumstats)
# load the matrix with the prior limits
data(limits)

# select a random simulation to act as target just to test the function
target <- sumstats[10 ,]
# we should remove the random simulation from the sumstats and params matrices
sumstats <- sumstats[-10, ]; params <- params[-10, ]

# parameter estimation for a single target
myabc <- singleABC(target = target, params = params, sumstats = sumstats,
limits = limits, tol = 0.01, method = "regression")

# check the fit of a summary statistic to the target
plot_stats(sumstat = sumstats, target = target, accepted = myabc$ss, index = 5)

# note that we performed parameter estimation for a single target
# because this function will only work when using a matrix