Adding scatterplot to existing scatterplot in plot_ly in R

Question

I wish to add scatterplots to an existing 3D scatterplot in R using plot_ly. I am generating data from Normal distribution of length 219. Then I want to plot and add scatter plots corresponding to the indices 80, 150 and 195 of the variables on the same graph with a different color (red in my case).

I used the following codes -

library(plotly)
index <- c(80, 150, 195)
set.seed(4991)
data1 <- rnorm(219,9,1.5)
data2 <- rnorm(219,2,1)
timeline <- 1:length(data1)

plot_ly(x=data1, y=data2, z=timeline, type="scatter3d", mode="markers")%>%
  layout(scene = list(xaxis = list(title = 'cases per day'),
                      yaxis = list(title = 'deaths per day'),
                      zaxis = list(title = 'observation #')))%>%
  add_trace(x=data1[index],y=data2[index],z=timeline[index], type="scatter3d", mode = "markers", marker = list(size = 5, color = "red", symbol = 104))

Snapshot of the output looks like this -

However if I use the color=timeline option in the plot_ly section,

plot_ly(x=data1, y=data2, z=timeline, type="scatter3d", mode="markers",color=timeline)%>%
  layout(scene = list(xaxis = list(title = 'cases per day'),
                      yaxis = list(title = 'deaths per day'),
                      zaxis = list(title = 'observation #')))%>%
  add_trace(x=data1[index],y=data2[index],z=timeline[index], type="scatter3d", mode = "markers", marker = list(size = 5, color = "red", symbol = 104))

I get the following error

Error:
! Tibble columns must have compatible sizes.
• Size 3: Columns `x`, `y`, and `z`.
• Size 219: Column `color`.
ℹ Only values of size one are recycled.
Run `rlang::last_error()` to see where the error occurred.

I want to plot the 3D scatterplot with color=timeline option and then add the scatterplot in red. Any help is appreciated.

Kat · Accepted Answer · 2022-08-10 12:42:46Z

1

The variable timeline is all unique values, which doesn't align with your desire to have the three values colored. What you need is a grouping variable (i.e., yes or no, a or b, etc.)

I made a control.

timeline1 <- rep("A", length(data1))
timeline1[index] <- "B"
summary(timeline1 %>% as.factor())
#   A   B 
# 216   3

Then I made my graph. One trace- with specific colors designated. I used Plotly's blue to keep it consistent with your question.

# '#1f77b4' is the Plotly blue (muted blue)
plot_ly(x = data1, y = data2, z = timeline, type = "scatter3d", mode = "markers",
        color = timeline1, colors = setNames(c('#1f77b4', "red"), nm = c("A", "B"))) %>%
  layout(scene = list(xaxis = list(title = 'cases per day'),
                      yaxis = list(title = 'deaths per day'),
                      zaxis = list(title = 'observation #')))

answered Aug 10, 2022 at 12:42

Kat

19k3 gold badges24 silver badges63 bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

Arnoneel Sinha Over a year ago

Great. That works. I was wondering if I plot my observations with color as per the observation list, as the following code - x1 <- rnorm(50,2,2) x2 <- rnorm(10,3.5,2) mydata <- data.frame(x1,x2) plot_ly(mydata, x = ~x1, y = ~x2, z = 1:length(x1), color = 1:length(x1)) How do I overlap the three dots on this 3D graph?

Kat Over a year ago

The only information I have to go on is the index values you provided. You only have 50 points in this version of the graph, so none of them are on it. However, if you wanted to highlight indices 10, 23, and 38, create data: d2 <- data.frame(x = x1[index], y = rep(x2, 5)[index], z = index) plot it: p2 <- plot_ly(d2, x = ~x, y = ~y, z = ~z, color = I("red"), type = "scatter3d", mode = "markers", showlegend = F) and smash it (plot in your comment is p1): subplot(p1, p2).

Arnoneel Sinha Over a year ago

Works perfectly. Got exactly what I wanted.

Collectives™ on Stack Overflow

Adding scatterplot to existing scatterplot in plot_ly in R

1 Answer 1

3 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

3 Comments

Your Answer

Sign up or log in

Post as a guest

Related