Generate a filled contour plot for DLNM output with custom color levels

This function is a modified version of dlnm::plot.crosspred() from the dlnm package. It generates a filled contour plot to visualize the estimated exposure-lag-response association from a distributed lag non-linear model (DLNM) object of class "crosspred".

Usage

plot_contour_dlnm(
  x,
  z_range = NULL,
  nlevels = 20,
  color1 = "blue",
  color2 = "red",
  color3 = "white",
  ci_level = x$ci.level,
  exp = NULL,
  ...
)

Arguments

x: An object of class "crosspred", typically generated using crosspred.
z_range: Optional numeric vector specifying the range of values for the contour levels. If NULL, the range defaults to the same automatic setting used in dlnm::plot.crosspred().
nlevels: Integer. The number of contour levels to display. Default is 20.
color1: Character string. Color used for negative (below reference) effects. Default is "blue".
color2: Character string. Color used for positive (above reference) effects. Default is "red".
color3: Character string. Color used for the reference effect level. Default is "white".
ci_level: Numeric. A value between 0 and 1 specifying the confidence level. Defaults to the ci_level stored in the x object.
exp: Logical. Indicating whether to exponentiate the estimates (e.g., for relative risks). If NULL, exponentiation is automatically applied for log or logit link models.
...: Additional graphical parameters passed to graphics::filled.contour().

Value

A filled contour plot.

Details

This function builds on dlnm::plot.crosspred() from the dlnm package, adding the ability to customize z-scale color levels via z_range, nlevels, and color*, which are set automatically in the original implementation.

Generate a contour plot of predicted effects over the predictor-lag space. Additional graphical options like plot.title, plot.axes, and key.title can be provided to adjust titles and axis/key labels. When z_range is not provided, the function produces a contour plot identical to that returned by dlnm::plot.crosspred().

References

Gasparrini A. “Distributed lag linear and non-linear models in R: the package dlnm.” Journal of Statistical Software. 2011; 43(8):1-20. https://doi.org/10.18637/jss.v043.i08.

Examples

# Fit a model and generate "crosspred" object
library(dlnm)
#> This is dlnm 2.4.10. For details: help(dlnm) and vignette('dlnmOverview').
library(splines)

cb.pm <- crossbasis(chicagoNMMAPS$pm10,
                    lag = 7,
                    argvar = list(fun = "lin"),
                    arglag = list(fun = "strata"))

varknots <- equalknots(chicagoNMMAPS$temp, fun = "bs", df = 5, degree = 2)
lagknots <- logknots(30, 3)
cb.temp <- crossbasis(chicagoNMMAPS$temp,
                      lag = 30,
                      argvar = list(fun = "bs", knots = varknots),
                      arglag = list(knots = lagknots))

model <- glm(death ~ cb.pm + cb.temp + ns(time, 7*14) + dow,
             family = quasipoisson(),
             data = chicagoNMMAPS)

pred.pm <- crosspred(cb.temp, model, cen = 21, by = 1)

# When `z_range` is not provided, the function produces a contour plot
# identical to that returned by `dlnm::plot.crosspred()`.
plot_contour_dlnm(pred.pm,
                  xlab = "Temperature",
                  key.title = title("RR"),
                  plot.title = title("Contour plot",
                                     xlab = "Temperature",
                                     ylab = "Lag"))


# Change `z_range` and `nlevels` to get user-defined color levels.
plot_contour_dlnm(pred.pm,
                  z_range = c(0.85, 1.4),
                  nlevels = 30,
                  xlab = "Temperature",
                  key.title = title("RR"),
                  plot.title = title("Contour plot",
                                     xlab = "Temperature",
                                     ylab = "Lag"))


# Customize colors
plot_contour_dlnm(pred.pm,
                  z_range = c(0.85, 1.4),
                  nlevels = 30,
                  color1 = "darkgreen",
                  xlab = "Temperature",
                  key.title = title("RR"),
                  plot.title = title("Contour plot",
                                     xlab = "Temperature",
                                     ylab = "Lag"))