ppplot: Probability-Probability Plot of FCS2 Model Fit
In aquaMetrics/fcs2: Fisheries Classification Scheme 2 For SNIFFER

Description Usage Arguments Details Value Warning See Also

Produces a Probability-Probability (P-P) plot of the fitted total catch that can be used to assess the model fit.

ppplot(
  fit,
  dataFrame,
  subset = 1:nrow(dataFrame),
  na.action,
  n.sims = 100,
  ciprob = 0.95,
  title = "",
  addCI = TRUE,
  progressBar = TRUE,
  mu,
  rho,
  seed = NULL
)

`fit`	an `"fcs2Fit"` object, usually returned by `fcs2FitModel`.
`dataFrame`	a data frame with surveys as rows and variables as columns. It should contain all variables required by `fit`. This is usually the same data frame used to create `fit`.
`subset`	an optional vector specifying a subset of surveys to be used to create the P-P plot.
`na.action`	a function which indicates what should happen when the data contain missing values (`NA`s). The default is set by the `na.action` setting of `options` and this is usually set to `na.omit`. This setting removes surveys that contain missing data in any required variables. A vector indicating the rows that were removed can be extracted from the returned object using `na.action.fcs2Fit`. Alternatively, `na.pass` can be used to ignore missing values (where possible) or `na.fail` can be given to signal an error if missing values are found.
`n.sims`	the number of datasets to simulate from the fitted model to generate the confidence interval if `addCI` is `TRUE`.
`ciprob`	the desired limits for the confidence interval added if `addCI` is `TRUE`. The default is `0.95` representing a 95% interval.
`title`	an optional title for the plot.
`addCI`	whether to add a confidence interval to the plot (default is `TRUE`).
`progressBar`	whether to show a progress bar when calculating the confidence interval, since this can take some time.
`mu`	a matrix of abundance samples can optionally be provided, as generated from `abundance` using the same `fit`, `dataFrame` and `subset`.
`rho`	a matrix of prevalence samples can optionally be provided, as generated from `prevalence` using the same `fit`, `dataFrame` and `subset`.
`seed`	set random seed to allow repeatable results.

If addCI is TRUE, a confidence interval is found by Monte Carlo simulation. n.sims datasets of fish catches are simulated from the fitted model, each using the same observed covariates, and the P-P plot line is found for each. The confidence interval then represents the extent covered by the central ciprob proportion of these.

If the observed data is typical of data simulated from the model, we would expect the observed P-P plot line to fall within the confidence interval ciprob * 100% of the time.

If addCI is TRUE, the proportion of observed P-P plot points within the confidence limits is printed to screen and invisibly returned.

Generating the confidence interval can take some time, especially if n.sims is large.

plot.fcs2Fit

aquaMetrics/fcs2 documentation built on Aug. 21, 2021, 12:55 p.m.