timeSerPlot: Plot time series of estimated _b_ with confidence intervals

In andrew-edwards/sizeSpectra: Fitting Size Spectra to Ecological Data Using Maximum Likelihood

Plot time series of estimated b with confidence intervals

Description

Plot time series of estimated b with confidence intervals, for data that are analysed year-by-year by a single method. And then fit a linear regression with its confidence intervals (or not if doRegression = FALSE). This will get called eight times to produce a comparison figure of the methods.

Usage

timeSerPlot(
  bForYears,
  legName,
  method,
  weightReg = FALSE,
  bCol = "black",
  pchVal = 20,
  cexVal = 1,
  confCol = "black",
  confThick = 1,
  xLim = NULL,
  yLim = NULL,
  xLab = "",
  yLab = expression(paste("Estimate of ", italic(b)), sep = ""),
  xTicksSmallInc = NULL,
  xTicksSmallTck = 0.01,
  yLabels = TRUE,
  yTicksSmallInc = NULL,
  yTicksSmallTck = 0.01,
  legPos = "topleft",
  newPlot = TRUE,
  regPlot = TRUE,
  regColNotSig = "darkgrey",
  regColSig = "red",
  legExtra = NULL,
  legExtraPos = "topleft",
  legExtraCol = "",
  insetVal = c(-0.08, -0.06),
  xJitter = 0,
  doRegression = TRUE
)

Arguments

bForYears	dataframe with columns Year, Method (optional), b (the estimate of b), confMin and confMax (the 95\ confidence limits) and stdError (the standard error of the estimate of b).
legName	legend name for that panel
method	method used to obtain the inputted estimates of b
weightReg	TRUE if doing weighted regression (using standard errors) or FALSE to not do weighted regression.
bCol	colour for points for b
pchVal	pch for points for b
cexVal	size of points for b
confCol	colour for confidence intervals for b
confThick	thickness of vertical line for confidence intervals
xLim	x-axis limits
yLim	y-axis limits
xLab	label for x-axis
yLab	label for y-axis
xTicksSmallInc	increments for where to have small (unlabelled) tickmarks on x-axis
xTicksSmallTck	tick length for small (unlabelled) tickmarks on x-axis
yLabels	whether or not to label main tickmarks on y-axis
yTicksSmallInc	increments for where to have small (unlabelled) tickmarks on y-axis
yTicksSmallTck	tick length for small (unlabelled) tickmarks on y-axis
legPos	legend position
newPlot	TRUE to create a new plot, FALSE to add to existing
regPlot	TRUE to plot the regression line and conf intervals
regColNotSig	colour for regression line (and its confidence intervals) if the trend is not significant
regColSig	colour for regression line (and its confidence intervals) if the trend is significant
legExtra	extra manually-specified legend (e.g. to distinguish two sets of results)
legExtraPos	position for extra manually-specified legend
legExtraCol	colours (vector) for extra manually-specified legend
insetVal	inset shift for naming the panel
xJitter	value to jitter the x-values by (for comparison plot the confidence intervals overlap)
doRegression	whether to calculate a regression for the time series of estimates

Value

if doRegression is TRUE (default) then return dataframe of just one row with columns (else return nothing):

• Method: method used

• Low: lower bound of 95\

• Trend: gradient of regression fit

• High: upper bound of 95\

• p: p-value of regression fit

• Rsquared: r-squared of regression fit

• adjRsquared: adjusted r-squared of regression fit

Author(s)

Andrew Edwards

andrew-edwards/sizeSpectra documentation built on June 28, 2023, 7:09 p.m.