R/draw-gam.R

In gavinsimpson/gratia: Graceful 'ggplot'-Based Graphics and Other Functions for GAMs Fitted Using 'mgcv'

#' Plot estimated smooths from a fitted GAM
#'
#' Plots estimated smooths from a fitted GAM model in a similar way to
#' `mgcv::plot.gam()` but instead of using base graphics, [ggplot2::ggplot()]
#' is used instead.
#'
#' @param object a fitted GAM, the result of a call to [mgcv::gam()].
#' @param data a optional data frame that may or may not be used? FIXME!
#' @param select character, logical, or numeric; which smooths to plot. If
#'   `NULL`, the default, then all model smooths are drawn. Numeric `select`
#'   indexes the smooths in the order they are specified in the formula and
#'   stored in `object`. Character `select` matches the labels for smooths
#'   as shown for example in the output from `summary(object)`. Logical
#'   `select` operates as per numeric `select` in the order that smooths are
#'   stored.
#' @param parametric logical; plot parametric terms also? Note that `select` is
#'   used for selecting which smooths to plot. The `terms` argument is used to
#'   select which parametric effects are plotted. The default, as with
#'   [mgcv::plot.gam()], is to not draw parametric effects.
#' @param terms character; which model parametric terms should be drawn? The
#'   Default of `NULL` will plot all parametric terms that can be drawn.
#' @param residuals logical; should partial residuals for a smooth be drawn?
#'   Ignored for anything but a simple univariate smooth.
#' @param scales character; should all univariate smooths be plotted with the
#'   same y-axis scale? If `scales = "free"`, the default, each univariate
#'   smooth has its own y-axis scale. If `scales = "fixed"`, a common y axis
#'   scale is used for all univariate smooths.
#'
#'   Currently does not affect the y-axis scale of plots of the parametric
#'   terms.
#' @param constant numeric; a constant to add to the estimated values of the
#'   smooth. `constant`, if supplied, will be added to the estimated value
#'   before the confidence band is computed.
#' @param fun function; a function that will be applied to the estimated values
#'   and confidence interval before plotting. Can be a function or the name of a
#'   function. Function `fun` will be applied after adding any `constant`, if
#'   provided.
#' @param ci_level numeric between 0 and 1; the coverage of credible interval.
#' @param n numeric; the number of points over the range of the covariate at
#'   which to evaluate the smooth.
#' @param n_3d numeric; the number of new observations to generate for the third
#'   dimension of a 3D smooth.
#' @param n_4d numeric; the number of new observations to generate for the
#'   dimensions higher than 2 (!) of a *k*D smooth (*k* >= 4). For example, if
#'   the smooth is a 4D smooth, each of dimensions 3 and 4 will get `n_4d`
#'   new observations.
#' @param unconditional logical; should confidence intervals include the
#'   uncertainty due to smoothness selection? If `TRUE`, the corrected Bayesian
#'   covariance matrix will be used.
#' @param overall_uncertainty logical; should the uncertainty in the model
#'  constant term be included in the standard error of the evaluate values of
#'  the smooth?
#' @param dist numeric; if greater than 0, this is used to determine when
#'   a location is too far from data to be plotted when plotting 2-D smooths.
#'   The data are scaled into the unit square before deciding what to exclude,
#'   and `dist` is a distance within the unit square. See
#'   [mgcv::exclude.too.far()] for further details.
#' @param rug logical; draw a rug plot at the bottom of each plot for 1-D
#'   smooths or plot locations of data for higher dimensions.
#' @param contour logical; should contours be draw on the plot using
#'   [ggplot2::geom_contour()].
#' @param grouped_by logical; should factor by smooths be drawn as one panel
#'   per level of the factor (`FALSE`, the default), or should the individual
#'   smooths be combined into a single panel containing all levels (`TRUE`)?
#' @param ci_alpha numeric; alpha transparency for confidence or simultaneous
#'   interval.
#' @param ci_col colour specification for the confidence/credible intervals
#'   band. Affects the fill of the interval.
#' @param smooth_col colour specification for the smooth line.
#' @param resid_col colour specification for the partial residuals.
#' @param contour_col colour specification for contour lines.
#' @param n_contour numeric; the number of contour bins. Will result in
#'   `n_contour - 1` contour lines being drawn. See [ggplot2::geom_contour()].
#' @param partial_match logical; should smooths be selected by partial matches
#'   with `select`? If `TRUE`, `select` can only be a single string to match
#'   against.
#' @param discrete_colour a suitable colour scale to be used when plotting
#'   discrete variables.
#' @param discrete_fill a suitable fill scale to be used when plotting
#'   discrete variables.
#' @param continuous_colour a suitable colour scale to be used when plotting
#'   continuous variables.
#' @param continuous_fill a suitable fill scale to be used when plotting
#'   continuous variables.
#' @param position Position adjustment, either as a string, or the result of a
#'   call to a position adjustment function.
#' @param angle numeric; the angle at which the x axis tick labels are to be
#'   drawn passed to the `angle` argument of [ggplot2::guide_axis()].
#' @param ncol,nrow numeric; the numbers of rows and columns over which to
#'   spread the plots
#' @param guides character; one of `"keep"` (the default), `"collect"`, or
#'   `"auto"`. Passed to [patchwork::plot_layout()]
#' @param widths,heights The relative widths and heights of each column and
#'   row in the grid. Will get repeated to match the dimensions of the grid. If
#'   there is more than 1 plot and `widths = NULL`, the value of `widths` will
#'   be set internally to `widths = 1` to accommodate plots of smooths that
#'   use a fixed aspect ratio.
#' @param crs the coordinate reference system (CRS) to use for the plot. All
#'   data will be projected into this CRS. See [ggplot2::coord_sf()] for
#'   details.
#' @param default_crs the coordinate reference system (CRS) to use for the
#'   non-sf layers in the plot. If left at the default `NULL`, the CRS used is
#'   4326 (WGS84), which is appropriate for spline-on-the-sphere smooths, which
#'   are parameterized in terms of latitude and longitude as coordinates. See
#'   [ggplot2::coord_sf()] for more details.
#' @param lims_method character; affects how the axis limits are determined. See
#'   [ggplot2::coord_sf()]. Be careful; in testing of some examples, changing
#'   this to `"orthogonal"` for example with the chlorophyll-a example from
#'   Simon Wood's GAM book quickly used up all the RAM in my test system and the
#'   OS killed R. This could be incorrect usage on my part; right now the grid
#'   of points at which SOS smooths are evaluated (if not supplied by the user)
#'   can produce invalid coordinates for the corners of tiles as the grid is
#'   generated for tile centres without respect to the spacing of those tiles.
#' @param wrap logical; wrap plots as a patchwork? If \code{FALSE}, a list of
#'   ggplot objects is returned, 1 per term plotted.
#' @param envir an environment to look up the data within.
#' @param ... additional arguments passed to [patchwork::wrap_plots()].
#'
#' @note Internally, plots of each smooth are created using [ggplot2::ggplot()]
#'   and composed into a single plot using [patchwork::wrap_plots()]. As a
#'   result, it is not possible to use `+` to add to the plots in the way one
#'   might typically work with `ggplot()` plots. Instead, use the `&` operator;
#'   see the examples.
#'
#' @return The object returned is created by [patchwork::wrap_plots()].
#'
#' @author Gavin L. Simpson
#'
#' @importFrom ggplot2 scale_colour_discrete scale_colour_continuous
#'   scale_fill_distiller
#' @importFrom patchwork wrap_plots
#' @importFrom dplyr mutate rowwise %>% ungroup left_join group_split summarise
#' @importFrom purrr pluck map_lgl
#' @importFrom rlang expr_label
#' @importFrom utils packageVersion getFromNamespace
#' @importFrom stringr str_split_fixed
#' @export
#'
#' @examples
#' load_mgcv()
#'
#' # simulate some data
#' df1 <- data_sim("eg1", n = 400, dist = "normal", scale = 2, seed = 2)
#' # fit GAM
#' m1 <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df1, method = "REML")
#'
#' # plot all smooths
#' draw(m1)
#'
#' # can add partial residuals
#' draw(m1, residuals = TRUE)
#'
#' df2 <- data_sim("eg2", n = 1000, dist = "normal", scale = 1, seed = 2)
#' m2 <- gam(y ~ s(x, z, k = 40), data = df2, method = "REML")
#' draw(m2, contour = FALSE, n = 50)
#'
#' # See https://gavinsimpson.github.io/gratia/articles/custom-plotting.html
#' # for more examples and for details on how to modify the theme of all the
#' # plots produced by draw(). To modify all panels, for example to change the
#' # theme, use the & operator
`draw.gam` <- function(
    object,
    data = NULL,
    select = NULL,
    parametric = FALSE,
    terms = NULL,
    residuals = FALSE,
    scales = c("free", "fixed"),
    ci_level = 0.95,
    n = 100,
    n_3d = 16,
    n_4d = 4,
    unconditional = FALSE,
    overall_uncertainty = TRUE,
    constant = NULL,
    fun = NULL,
    dist = 0.1,
    rug = TRUE,
    contour = TRUE,
    grouped_by = FALSE,
    ci_alpha = 0.2,
    ci_col = "black",
    smooth_col = "black",
    resid_col = "steelblue3",
    contour_col = "black",
    n_contour = NULL,
    partial_match = FALSE,
    discrete_colour = NULL,
    discrete_fill = NULL,
    continuous_colour = NULL,
    continuous_fill = NULL,
    position = "identity",
    angle = NULL,
    ncol = NULL, nrow = NULL,
    guides = "keep", widths = NULL, heights = NULL,
    crs = NULL,
    default_crs = NULL,
    lims_method = "cross",
    wrap = TRUE,
    envir = environment(formula(object)),
    ...
  ) {
  # fixed or free?
  scales <- match.arg(scales)

  # fix up default scales
  # if (is.null(discrete_colour)) {
  #    discrete_colour <- scale_colour_discrete()
  # }
  if (is.null(continuous_colour)) {
    continuous_colour <- scale_colour_continuous()
  }
  if (is.null(continuous_fill)) {
    continuous_fill <- scale_fill_distiller(palette = "RdBu", type = "div")
  }

  # if not using select, set parametric TRUE if not set to FALSE
  if (!is.null(select)) {
    if (is.null(parametric)) {
      parametric <- FALSE
    }
  } else {
    if (is.null(parametric)) {
      parametric <- TRUE
    }
  }

  # sort out n_3d and n_4d. If these are `NULL` then do sensible thing at set
  # them small. Default is 3 for n_4d and for 12 for n_3d, but if we have a kD
  # smooth (k >= 4) we want to

  S <- smooths(object) # vector of smooth labels - "s(x)"

  # select smooths
  select <-
    check_user_select_smooths(
      smooths = S, select = select,
      partial_match = partial_match,
      model_name = expr_label(substitute(object))
    )

  # this is needed for the parametric terms below
  sm_plts <- NULL
  ylims <- NULL

  # do we have any smooths to plot?
  if (length(select) > 0L) {
    # evaluate all requested smooths
    sm_eval <- smooth_estimates(object,
      select = S[select],
      n = n,
      n_3d = n_3d,
      n_4d = n_4d,
      data = data,
      unconditional = unconditional,
      overall_uncertainty = overall_uncertainty,
      dist = dist,
      unnest = FALSE
    )

    # grab tensor term order if present, if not it is NULL & that's OK
    tensor_term_order <- attr(sm_eval, "tensor_term_order")

    # add confidence interval
    sm_eval <- sm_eval %>%
      rowwise() %>%
      mutate(data = list(add_confint(.data$data, coverage = ci_level))) %>%
      ungroup()

    # Take the range of the smooths & their confidence intervals now
    # before we put rug and residuals on
    if (utils::packageVersion("dplyr") > "1.0.10") {
      sm_rng <- sm_eval |>
        rowwise() |>
        utils::getFromNamespace("reframe", "dplyr")(rng =
          range(c(data$.estimate, data$.lower_ci, data$.upper_ci))) |>
        pluck("rng")
    } else {
      sm_rng <- sm_eval |>
        rowwise() |>
        summarise(rng = range(c(
          data$.estimate, data$.lower_ci,
          data$.upper_ci
        ))) |>
        pluck("rng")
    }

    # Add partial residuals if requested - by default they are
    # At the end of this, sm_eval will have a new list column containing the
    # partial residuals, `partial_residual`
    p_resids_rng <- NULL
    if (isTRUE(residuals)) {
      if (is.null(residuals(object)) || is.null(weights(object))) {
        residuals <- FALSE
      } else {
        # get residuals in a suitable format
        p_resids <- nested_partial_residuals(object, terms = S[select])

        # compute the range of residuals for each smooth
        # p_resids_rng <- p_resids |>
        #     rowwise() |>
        #     dplyr::reframe(rng =
        #         range(.data$partial_residual$partial_residual)) |>
        #     pluck("rng")
        if (utils::packageVersion("dplyr") > "1.0.10") {
          p_resids_rng <- p_resids |>
            rowwise() |>
            utils::getFromNamespace("reframe", "dplyr")(
              rng = range(.data$partial_residual$partial_residual)) |>
            pluck("rng")
        } else {
          p_resids_rng <- p_resids |>
            rowwise() |>
            summarise(
              rng =
                range(.data$partial_residual$partial_residual)
            ) |>
            pluck("rng")
        }
        # merge with the evaluated smooth
        sm_eval <- suppress_matches_multiple_warning(
          left_join(sm_eval, p_resids, by = ".smooth")
        )
      }
    }

    # add rug data?
    if (isTRUE(rug)) {
      # get rug data in a suitable format
      rug_data <- nested_rug_values(object, terms = S[select])

      # merge with the evaluated smooth
      sm_eval <- suppress_matches_multiple_warning(
        left_join(sm_eval, rug_data, by = ".smooth")
      )
    }

    # need to figure out scales if "fixed"
    if (isTRUE(identical(scales, "fixed"))) {
      ylims <- range(sm_rng, p_resids_rng)
    }

    # draw smooths
    sm_l <- if (isTRUE(grouped_by)) {
      sm_levels <- unique(sm_eval$.smooth)
      levs <- unique(str_split_fixed(sm_eval$.smooth, ":", n = 2)[, 1])
      sm_l <- sm_eval |>
        mutate(
          ..smooth.. = factor(.data$.smooth, levels = S[select]),
          .term = str_split_fixed(.data$.smooth, ":", n = 2)[, 1],
          ..by.. = if_else(is.na(.data$.by), "..no_level..", .data$.by)
        ) |>
        relocate(".term", .before = 1L)
      grp_by_levs <- unique(sm_l$"..by..")
      sm_l <- sm_l |>
        group_split(factor(.data$.term, levels = sm_levels),
          factor(.data$"..by..", levels = grp_by_levs))
      # sometimes the steps to get the order right above don't work
      sm_l_levs <- vapply(sm_l, \(x) unique(x$.term), character(1L))
      if (!identical(unique(sm_l_levs), levs)) {
        names(sm_l) <- sm_l_levs
        sm_l <- sm_l[levs]
        names(sm_l) <- NULL
      }
      sm_l
    } else {
      # the factor is to reorder to way the smooths entered the model
      group_split(sm_eval, factor(.data$.smooth, levels = S[select]))
    }
    sm_plts <- map(sm_l,
      draw_smooth_estimates,
      constant = constant,
      fun = fun,
      contour = contour,
      contour_col = contour_col,
      n_contour = n_contour,
      ci_alpha = ci_alpha,
      ci_col = ci_col,
      smooth_col = smooth_col,
      resid_col = resid_col,
      partial_match = partial_match,
      discrete_colour = discrete_colour,
      discrete_fill = discrete_fill,
      continuous_colour = continuous_colour,
      continuous_fill = continuous_fill,
      angle = angle,
      ylim = ylims,
      crs = crs,
      default_crs = default_crs,
      lims_method = lims_method,
      tensor_term_order = tensor_term_order
    )
  } # end stuff for smooths...

  # Are we plotting parametric effects too?
  if (isTRUE(parametric)) {
    if (length(parametric_terms(object)) == 0L) {
      message("The model contains no parametric terms")
      parametric <- FALSE
    } else {
      para <- parametric_effects(object,
        select = terms, data = data,
        unconditional = unconditional,
        unnest = TRUE, ci_level = ci_level, envir = envir
      )
      # Add CI
      # crit <- coverage_normal(ci_level)
      # object <- mutate(para,
      #    .lower_ci = .data$.partial - (crit * .data$.se),
      #    .upper_ci = .data$.partial + (crit * .data$.se))
      object <- para |> add_confint(coverage = ci_level)
      # need to alter the ylim if scales are fixed
      if (isTRUE(identical(scales, "fixed"))) {
        ylims <- range(
          ylims, object$.partial, object$.upper_ci,
          object$.lower_ci
        )
      }

      para_plts <- para %>%
        group_by(.data$term) %>%
        group_map(
          .keep = TRUE,
          .f = ~ draw_parametric_effect(.x,
            ci_level = ci_level,
            ci_col = ci_col,
            ci_alpha = ci_alpha,
            line_col = smooth_col,
            constant = constant,
            fun = fun,
            rug = rug,
            position = position,
            angle = angle,
            ylim = ylims
          )
        )
    }
  }

  if (isTRUE(parametric)) {
    sm_plts <- append(sm_plts, para_plts)
  }

  # filter out NULLs as those are types of smooths we can't plot (yet)
  no_plot <- map_lgl(sm_plts, is.null)
  sm_plts <- sm_plts[!no_plot]

  if (all(no_plot)) {
    message("Unable to draw any of the model terms.")
    return(invisible())
  }

  # return
  n_plots <- length(sm_plts)
  if (is.null(ncol) && is.null(nrow)) {
    ncol <- ceiling(sqrt(n_plots))
    nrow <- ceiling(n_plots / ncol)
  }
  if (n_plots > 1L && is.null(widths)) {
    # it doesn't matter about the widths if only one plot, but if we have
    # more than one plot and the user didn't change `widths`, then we will
    # force a value of 1 to give all plots the same relative width
    widths <- 1
  }

  if (wrap) {
    sm_plts <- wrap_plots(sm_plts,
      byrow = TRUE, ncol = ncol, nrow = nrow,
      guides = guides, widths = widths, heights = heights, ...
    )
  }

  # return
  sm_plts
}

#' @export
`draw.list` <- function(object, ...) {
  if (!is_gamm4(object)) {
    stop("Don't know how to draw a list")
  }
  draw(object$gam, ...)
}

#' @export
`draw.gamm` <- function(object, ...) {
  draw(object$gam, ...)
}

#' @export
`draw.scam` <- function(object, ..., overall_uncertainty = FALSE) {
  class(object) <- append(class(object), "gam", after = 1)
  draw.gam(object, overall_uncertainty = FALSE, ...)
}

# TODO: Remove after dplyr 1.1.0 is released and use `multiple = "all"` instead
suppress_matches_multiple_warning <- function(expr) {
  handler_matches_multiple <- function(cnd) {
    if (inherits(cnd, "dplyr_warning_join_matches_multiple")) {
      restart <- findRestart("muffleWarning")
      if (!is.null(restart)) {
        invokeRestart(restart)
      }
    }
  }
  withCallingHandlers(expr, warning = handler_matches_multiple)
}