Constant heatmap across animated frames using plotly

user_9034199 0
Viewed 198

I want to create an animated figure, where a heatmap works as the "background". That is, I want the heatmap to be constant across the frames.

This is my code so far:

library(plotly)
library(dplyr)

df1 <- data.frame(f = rep(1:5, c(rep(50, 5))), 
                  x = round(runif(250, 10,100),0),
                  y = round(runif(250, 10,100),0)) 


df2 <- expand.grid(x=seq(0,100,10), y=seq(0,100,10))
df2$col[df2$x>50] = "green"
df2$col[df2$x<=50] = "red"
df2$col[df2$x<=20] = "blue"



plot_ly() %>%
  plotly::add_markers(data = df1, y = ~y,x = ~x,frame=~f)%>%
  plotly::add_heatmap(data = df2,
                      x = ~x,
                      y = ~y,
                      z = as.numeric(factor(df2$col, ordered = T, levels = unique(df2$col))),
                      xgap = 0.5,
                      ygap = 0.5)

producing:

enter image description here

I can create the map and on frame 1 everything looks fine. However, when I move one frame ahead the heatmap vanishes. Anyone having a solution to this?

By the way, ggplotly Is not an options as the combination of plotly and heatmap are too slow.

plotly heatmap r
2 Answers

Have you tried repeating the df2 object by the number of frames? Although the heatmap is reconstructed resulting in a noticeable delay between frames, the below is a potential solution:

# repeat df2 five times
df2 = do.call("rbind", replicate(5, df2, simplify = FALSE))
df2$f = rep(1:5, each = nrow(df2)/5)

plot_ly() %>%
  plotly::add_markers(data = df1, y = ~y,x = ~x,frame=~f)%>%
  plotly::add_heatmap(data = df2,
                      x = ~x,
                      y = ~y,
                      frame=~f,
                      z = as.numeric(factor(df2$col, ordered = T, levels = unique(df2$col))),
                      xgap = 0.5,
                      ygap = 0.5)
user_149462670

You might want to consider shapes instead of heatmap.

library(plotly)
library(dplyr)

df1 <- data.frame(f = rep(1:5, c(rep(50, 5))), 
                  x = round(runif(250, 10,100),0),
                  y = round(runif(250, 10,100),0)) 

df2 <- expand.grid(x=seq(0,100,10), y=seq(0,100,10))
df2$col[df2$x>50] = "green"
df2$col[df2$x<=50] = "red"
df2$col[df2$x<=20] = "blue"

shapes <- apply(X = df2, MARGIN = 1, FUN = function(hm) { list(type = 'rect', fillcolor = hm['col'],opacity = 0.3, line = list(color = "white"), 
                                                   x0 = as.numeric(hm['x']) - 5, x1 = as.numeric(hm['x']) + 5, xref = "x",
                                                   y0 = as.numeric(hm['y']) - 5, y1 = as.numeric(hm['y']) + 5, yref = "y") })

plot_ly() %>%
    plotly::add_markers(data = df1, y = ~y,x = ~x,frame=~f) %>%
    layout(shapes = shapes)
user_104448550
Related