Plot (clustered) heatmaps using pretty ggplot theme.

`plot_heatmap` generates nice heatmaps of a data matrix (or frame) with ease. `plot_hclust_heatmap` provides additional functionality beyond `plot_heatmap` that allows for clustering the rows and columns of the heatmap via hierarchical clustering. `plot_cor_heatmap` generates nice (clustered) heatmaps for the feature correlation matrix of X. These heatmap functions also allow for row/column groups in the plotting display, colored axis text labels, and customization of the color/theme.

plot_heatmap(
  X,
  y_groups = NULL,
  x_groups = NULL,
  ytext_labels = rownames(X),
  xtext_labels = colnames(X),
  ytext_num = FALSE,
  xtext_num = FALSE,
  ytext_colors = NULL,
  xtext_colors = NULL,
  show_ytext = TRUE,
  show_xtext = TRUE,
  center = FALSE,
  scale = FALSE,
  z_range = NULL,
  text_size = 0,
  y_orient = "identity",
  size = 0,
  color_by_quantile = FALSE,
  n_quantiles = 5,
  color_scheme = "viridis",
  viridis_option = "C",
  theme_function = NULL,
  show_plot = FALSE,
  ...
)

plot_hclust_heatmap(
  X,
  y_groups = NULL,
  x_groups = NULL,
  ytext_labels = rownames(X),
  xtext_labels = colnames(X),
  ytext_num = FALSE,
  xtext_num = FALSE,
  ytext_colors = NULL,
  xtext_colors = NULL,
  show_ytext = TRUE,
  show_xtext = TRUE,
  clust_x = TRUE,
  clust_y = TRUE,
  clust_x_wi_group = TRUE,
  clust_y_wi_group = TRUE,
  dist_metric_x = "euclidean",
  dist_metric_y = "euclidean",
  dist_matrix_x = NULL,
  dist_matrix_y = NULL,
  linkage_x = "ward.D",
  linkage_y = "ward.D",
  center = FALSE,
  scale = FALSE,
  z_range = NULL,
  text_size = 0,
  size = 0,
  color_by_quantile = FALSE,
  n_quantiles = 5,
  color_scheme = "viridis",
  viridis_option = "C",
  theme_function = NULL,
  show_plot = FALSE,
  ...
)

plot_cor_heatmap(
  X,
  cor_type = "pearson",
  absolute_value = FALSE,
  xytext_labels = colnames(X),
  xytext_colors = NULL,
  show_ytext = TRUE,
  show_xtext = TRUE,
  clust = TRUE,
  linkage = "ward.D",
  z_range = c(-1, 1),
  text_size = 0,
  color_by_quantile = FALSE,
  n_quantiles = 5,
  color_scheme = "cor_temperature",
  viridis_option = "C",
  theme_function = NULL,
  show_plot = FALSE,
  ...
)

Arguments

X: Data matrix or data frame to use for heatmap.
y_groups: Data vector of group ids to use for grouping rows in heatmap.
x_groups: Data vector of group ids to use for grouping columns in heatmap.
ytext_labels: y-axis labels in heatmap.
xtext_labels: x-axis labels in heatmap.
ytext_num: Logical; whether or not y labels are numeric/continuous.
xtext_num: Logical; whether or not x labels are numeric/continuous.
ytext_colors: (Optional) Data vector to use for coloring y-axis text.
xtext_colors: (Optional) Data vector to use for coloring x-axis text.
show_ytext: Logical; whether or not to show y-axis text labels.
show_xtext: Logical; whether or not to show x-axis text labels.
center: Logical; whether or not to center columns of X.
scale: Logical; whether or not to scale columns of X.
z_range: Vector of length 2 with the min and max of the fill color range for heatmap. Used to set bounds for fill legend.
text_size: Numeric; size of text on heatmap. If text_size = 0 (default), no text is shown.
y_orient: One of "identity" or "ordered". If "identity", plots heatmap of X as if it is an image. If "ordered", plots first row of X at the bottom of the heatmap and the last row of X on top.
size: Size argument in [ggplot2::geom_tile()] to avoid white lines in continuous heatmap.
color_by_quantile: Logical; whether or not to use quantiles to construct fill color scale. Default is FALSE.
n_quantiles: Number of quantiles for color scheme. Used only if color_by_quantile = TRUE.
color_scheme: One of "viridis", "temperature", "cor_temperature", or a data vector of colors to use for the fill color scale.
viridis_option: Argument indicating viridis palette name. Only used if `color_scheme = "viridis"`.
theme_function: Function which adds theme() to ggplot object. If NULL, adds vthemes::theme_vmodern() to ggplot.
show_plot: Logical. Should this plot be printed? Default FALSE.
...: Other arguments to pass to vthemes::theme_vmodern() or theme_function().
clust_x: Logical; whether or not to cluster columns.
clust_y: Logical; whether or not to cluster rows.
clust_x_wi_group: Logical; whether or not cluster within x_groups.
clust_y_wi_group: Logical; whether or not cluster within y_groups.
dist_metric_x: Distance metric for clustering columns (see [stats::dist()]).
dist_metric_y: Distance metric for clustering rows (see [stats::dist()]).
dist_matrix_x: Distance matrix for clustering columns (optional). Must provide either dist_metric or dist_mat if clust_x = T.
dist_matrix_y: Distance matrix for clustering rows (optional). Must provide either dist_metric or dist_mat if clust_y = T.
linkage_x: Type of linkage for clustering columns (see [stats::hclust()]).
linkage_y: Type of linkage for clustering rows (see [stats::hclust()]).
cor_type: Type of correlation. Must be one of "pearson", "kendall", or "spearman"
absolute_value: Logical indicating whether or not to show absolute value of the correlation values in the correlation heatmap.
xytext_labels: x- and y-axis labels for correlation heatmap.
xytext_colors: (Optional) Data vector to use for coloring x- and y-axis text labels in correlation heatmap.
clust: Logical; whether or not to cluster columns and rows in correlation heatmap
linkage: Type of linkage for clustering rows and columns in correlation heatmap (see [stats::hclust()]).

Value

A ggplot object.

Examples

df <- as.data.frame(matrix(1:100, nrow = 10, byrow = 10))
plot_heatmap(df,  y_orient = "identity")

plot_heatmap(df,  y_orient = "ordered")

plot_hclust_heatmap(df)

plot_cor_heatmap(df)