Dear Cory, how to set conditions to filter the flows?

[peter308]

Dear Cory,

In the Geom_flow example. we can set max.y, and min.y to filter flows.
Do you know where I can set separate conditions for each flow?
Say I have 10 flows, flow 1 statisfy condition then I will show it, flow 2 violates the condition, then I will not show it, etc.

Can you sugget which part of the code I can modify to realize this? Tks.

with warm regards,
Peter

[corybrunson]

Hi @peter308, i think this should be done manually, by introducing a binary variable to the data that you can then use to set some flows to transparent and leave others alone, for example. I expect that trying to adapt the min.y and max.y parameters to vary by flow, or even by axis, would be more complex than the benefits would justify.

[peter308]

Dear Cory,

I didn't find a way to set the transparency of the flows. Are you suggesting me to set the transparency of the lodes?
But that will also change the transparency of flows that I want to keep.

Now I try to remove the flows I don't want by removing them from the dataframe.
Although the problem is solved and is not urgent right now, still, if you can let me know on how to change the transparency of the flows, it will be useful for me in the future. Tks.

With best regards,
Peter

[corybrunson]

Sure, here are two ways to do it: first based on frequencies in the original data (cohorts, i.e. alluvia), second based on the combined frequencies that define the flows. (I apologize—i'm having problems with {reprex} on my aging work computer and cannot use it to display executed code.)

library(ggalluvial)

# hide alluvia below survey-specific frequencies
vaccinations2 <- vaccinations |> 
  transform(
    hide = 
      (survey == "ms153_NSA" & freq < 25L) |
      (survey == "ms432_NSA" & freq < 100L)
  )
ggplot(
  vaccinations2,
  aes(x = survey, y = freq, alluvium = subject, stratum = response)
) +
  geom_flow(aes(fill = response, alpha = hide)) +
  geom_stratum(aes(fill = response)) +
  scale_alpha_manual(values = c(`FALSE` = 2/3, `TRUE` = 0)) +
  guides(alpha = "none")

# hide flows below survey-specific frequencies
ggplot(
  vaccinations,
  aes(x = survey, y = freq, alluvium = subject, stratum = response)
) +
  geom_flow(aes(
    fill = response,
    alpha = after_stat(( x == 1L & count < 50L ) | ( x == 2L & count < 200L))
  )) +
  geom_stratum(aes(fill = response)) +
  scale_alpha_manual(values = c(`FALSE` = 2/3, `TRUE` = 0)) +
  guides(alpha = "none")

[peter308]

Dear Cory,

Well noted with thanks. Thanks for your kindness and help.

with best regards,
Peter