# Plotting Univariate Mark-Specific Proportional Hazards Model Fits Using `ggplot` `ggplot`-style plotting for univariate marks. Point and interval estimates of the mark-specific treatment effect parameter specified by component `contrast` in `summary.sievePH` or `summary.kernel_sievePH` are plotted, together with scatter and box plots of the observed mark values by treatment. ```r ggplot_sieve( x, mark = NULL, tx = NULL, xlim = NULL, ylim = NULL, xtickAt = NULL, xtickLab = NULL, ytickAt = NULL, ytickLab = NULL, tickLabSize = 14, xlab = NULL, ylab = NULL, axisLabSize = 15, title = NULL, titleSize = 16, subtitle = NULL, subtitleSize = 10, txLab = c("Placebo", "Treatment"), txLabSize = 5, legendLabSize = 12, legendPosition = c(0.96, 1.08), legendJustification = c(1, 1), estLineSize = 1.6, ciLineSize = 1.2, boxplotWidth = 0.8, jitterFactor = 0.1, jitterSeed = 0, pointColor = c("blue", "red3"), pointSize = 1.7, bottomPlotMargin = c(-0.5, 0.3, 0, 0), topPlotMargin = c(0, 0.3, -0.12, 1.83), plotHeights = c(0.33, 0.67) ) ``` ## Arguments - `x`: an object returned by `summary.sievePH` or `summary.kernel_sievePH` - `mark`: a numeric vector specifying a univariate continuous mark. For subjects with a right-censored time-to-event, the value(s) in `mark` should be set to `NA`. - `tx`: a numeric vector indicating the treatment group (1 if treatment, 0 if placebo) - `xlim`: a numeric vector of length 2 specifying the x-axis range (`NULL` by default) - `ylim`: a numeric vector of length 2 specifying the y-axis range (`NULL` by default) - `xtickAt`: a numeric vector specifying the position of x-axis tickmarks (`NULL` by default) - `xtickLab`: a numeric vector specifying labels for tickmarks listed in `xtickAt`. If `NULL` (default), the labels are determined by `xtickAt`. - `ytickAt`: a numeric vector specifying the position of y-axis tickmarks (`NULL` by default) - `ytickLab`: a numeric vector specifying labels for tickmarks listed in `ytickAt`. If `NULL` (default), the labels are determined by `ytickAt`. - `tickLabSize`: a numeric value specifying the font size of tickmark labels along both axes in the bottom panel (`14` by default) - `xlab`: a character string specifying the x-axis label (`NULL` by default) - `ylab`: a character string specifying the y-axis label (`NULL` by default) - `axisLabSize`: a numeric value specifying the font size of both axis labels in the bottom panel (`15` by default) - `title`: a character string specifying the plot title (`NULL` by default) - `titleSize`: a numeric value specifying the font size of the plot title (`16` by default) - `subtitle`: a character string specifying the plot subtitle (`NULL` by default) - `subtitleSize`: a numeric value specifying the font size of the plot subtitle (`10` by default) - `txLab`: a character vector of length 2 specifying the placebo and treatment labels (in this order). The default labels are `placebo` and `treatment`. - `txLabSize`: a numeric value specifying the font size of labels `txLab` (`5` by default) - `legendLabSize`: a numeric value specifying the font size of legend labels in the bottom panel (`11` by default) - `legendPosition`: a numeric vector of length 2 specifying the position of the legend in the bottom panel (`c(0.96, 1.08)` by default), passed on to argument `legend.position` in `theme()` - `legendJustification`: a numeric vector of length 2 specifying the justification of the legend in the bottom panel (`c(1, 1)` by default), passed on to argument `legend.justification` in `theme()` - `estLineSize`: a numeric value specifying the line width for the point estimate of the mark-specific treatment effect (`1.6` by default) - `ciLineSize`: a numeric value specifying the line width for the confidence limits for the mark-specific treatment effect (`1.2` by default) - `boxplotWidth`: a numeric value specifying the width of each box in the box plot (`0.8`) by default - `jitterFactor`: a numeric value specifying the amount of vertical jitter (`0.1` by default) - `jitterSeed`: a numeric value setting the seed of R's random number generator for jitter in the scatter plot (`0` by default) - `pointColor`: a character vector of length 2 color-coding the placebo and treatment group (in this order) in the scatter plot (`c("blue", "red3")` by default) - `pointSize`: a numeric value specifying the size of data points in the scatter plot (`1.7` by default) - `bottomPlotMargin`: a numeric vector, using cm as the unit, passed on to argument `plot.margin` in `theme()` for the bottom panel (`c(-0.5, 0.3, 0, 0)` by default) - `topPlotMargin`: a numeric vector, using `"lines"` as the unit, passed on to argument `plot.margin` in `theme()` for the top panel (`c(0, 0.3, -0.12, 1.83)` by default) - `plotHeights`: a numeric vector specifying relative heights of the top and bottom panels (`c(0.33, 0.67)` by default) passed on to argument `heights` in `ggpubr::ggarrange()` ## Returns A `ggplot` object. ## Examples ```r n <- 200 tx <- rep(0:1, each=n/2) tm <- c(rexp(n/2, 0.2), rexp(n/2, 0.2 * exp(-0.4))) cens <- runif(n, 0, 15) eventTime <- pmin(tm, cens, 3) eventInd <- as.numeric(tm <= pmin(cens, 3)) mark <- ifelse(eventInd==1, c(rbeta(n/2, 2, 5), rbeta(n/2, 2, 2)), NA) markRng <- range(mark, na.rm=TRUE) # fit a model with a univariate mark using the sievePH method fit1 <- sievePH(eventTime, eventInd, mark, tx) sfit1 <- summary(fit1, markGrid=seq(markRng[1], markRng[2], length.out=10)) print(ggplot_sieve(sfit1, mark, tx)) # fit a model with a univariate mark using the kernel_sievePH method fit2 <- kernel_sievePH(eventTime, eventInd, mark, tx, tau = 3, tband = 0.5, hband = 0.3, nvgrid = 20, nboot = NULL) sfit2 <- summary(fit2) print(ggplot_sieve(sfit2, mark, tx)) ``` ## See Also `plot.summary.sievePH`, `sievePH`, `summary.sievePH`, `kernel_sievePH`, `summary.kernel_sievePH`

ggplot_sieve function