How to manually assign linetype to lines in plotly

Question

I have a plot like this:

library(plotly)
library(tidyr)
library(dplyr)

data.frame(date = seq(ymd('2020-01-01'), ymd('2020-05-31'), by = '1 day'), 
           sin = sin(seq(-pi, pi, length.out = 152)), 
           cos = cos(seq(-pi, pi, length.out = 152))) %>% 
  gather(k, v, -date) %>% 
  plot_ly() %>% 
  add_lines(x = ~date, y = ~v, linetype = ~k, color = I('black'))

The output looks like below:

What I'd like to do is to manually assign linetype to values of k variable, and for example, assign dashed line to cos and solid line to sin no matter of their order. Basically I'm looking for equivalent for ggplot2::scale_linetype_manual in plotly? How can I achieve this without adding new trace for each line? (I don't know the exact number of k values beforehand)

Answer 1

Similar to ggplot2 this can be achieved by using a named vector of linetypes like so:

library(plotly)
library(tidyr)
library(dplyr)
library(lubridate)

lty <- c(cos = "dash", sin = "solid")

data.frame(date = seq(ymd('2020-01-01'), ymd('2020-05-31'), by = '1 day'), 
           sin = sin(seq(-pi, pi, length.out = 152)), 
           cos = cos(seq(-pi, pi, length.out = 152))) %>% 
  gather(k, v, -date) %>% 
  plot_ly() %>% 
  add_lines(x = ~date, y = ~v, linetype = ~k, color = I('black'), linetypes = lty)