Playing with {rtoot}

I was curious if there’s a positive correlation between the total number of users on a server and how many followers I have from that server.

Include some packages…

library(rtoot)
library(stringr)
library(tidyverse)

── Attaching core tidyverse packages ──────────────────────────── tidyverse 2.0.0 ──
✔ dplyr     1.1.2     ✔ purrr     1.0.1
✔ forcats   1.0.0     ✔ readr     2.1.4
✔ ggplot2   3.4.2     ✔ tibble    3.2.1
✔ lubridate 1.9.2     ✔ tidyr     1.3.0── Conflicts ────────────────────────────────────────────── tidyverse_conflicts() ──
✖ dplyr::filter() masks stats::filter()
✖ dplyr::lag()    masks stats::lag()
ℹ Use the ]8;;http://conflicted.r-lib.org/conflicted package]8;; to force all conflicts to become errors

Get {rtoot} authorised to talk to my server:

auth_setup(instance = "tech.lgbt", type ="user")

I’ll need my ID: who am I?

acc <- search_accounts("@Andi@tech.lgbt")
acc |>
  select(id, acct, display_name) |>
  head(1)

It me!

whoami <- "109273348690338129"

There’s a handy function in {rtoot} for getting all followers; however, it doesn’t (or didn’t, end of 2022) support auto-pagination. After reading the friendly manual, here’s a workaround:

really_get_all_followers <- function(id, sure = "No!") {
  stopifnot(sure == "Yes, I know what I am doing")
  
  followers <- c()
  still_working <- TRUE
  max_id <- NULL
  
  while (still_working) {
    next_lot <- get_account_followers(id,
                                      max_id = max_id)
    followers <- bind_rows(followers, next_lot)
  
    attrs <- attr(next_lot, "headers")
    if ("max_id" %in% names(attrs))
      max_id <- attrs$max_id
    else
      still_working <- FALSE
  }
  
  followers
}

Get my followers:

my_followers <- really_get_all_followers(
  whoami,
  sure = "Yes, I know what I am doing"
)

This number is correct: it worked!

nrow(my_followers)

[1] 494

What servers are they from?

get_servers <- function(followers) {
  servers <- followers$acct |> str_split_fixed("@", 2)
  servers[,2]
}

followers_servers <- my_followers |>
  mutate(server = get_servers(my_followers)) |>
  mutate(server = ifelse(server == "", "tech.lgbt", server))

Here are the counts:

server_count <- followers_servers |>
  group_by(server) |>
  summarise(n = n()) |>
  arrange(desc(n))

server_count

Check everything adds up:

server_count$n |> sum()

[1] 494

So far so good.

Next up, how many users are there on each of those servers? Note the exception handling…

get_user_count <- function(server) {
  res <- NA
  
  # This will catch problems like missing servers
  tryCatch(
    res <- get_instance_general(server)$stats$user_count,
    error = function(e) {
        cat("***")
        cat(server)
        cat("***")
        cat("\n")
        print(e)
        cat("\n")
      }
  )
  ifelse(is.numeric(res), ifelse(length(res) == 1, res, NA), NA)
}

server_count$server_user_n <- map(server_count$server, get_user_count)

***bbs.crumplab.com***
<simpleError in curl::curl_fetch_memory(url, handle = handle): SSL peer certificate or SSH remote key was not OK: [bbs.crumplab.com] schannel: SEC_E_UNTRUSTED_ROOT (0x80090325) - The certificate chain was issued by an authority that is not trusted.>

***bv.umbrellix.org***
<simpleError: something went wrong. Status code: 503>

***bytebuilders.uk***
<simpleError in curl::curl_fetch_memory(url, handle = handle): SSL peer certificate or SSH remote key was not OK: [bytebuilders.uk] schannel: SEC_E_UNTRUSTED_ROOT (0x80090325) - The certificate chain was issued by an authority that is not trusted.>

***calckey.social***
<simpleError: something went wrong. Status code: 500>

***fedi.astrid.tech***
<simpleError in curl::curl_fetch_memory(url, handle = handle): Timeout was reached: [fedi.astrid.tech] Connection timeout after 10012 ms>

***firefish.social***
<simpleError: something went wrong. Status code: 500>

***iscurrently.live***
<simpleError: something went wrong. Status code: 522>

No encoding supplied: defaulting to UTF-8.

***mythago.space***
<simpleError in curl::curl_fetch_memory(url, handle = handle): SSL peer certificate or SSH remote key was not OK: [mythago.space] schannel: SEC_E_UNTRUSTED_ROOT (0x80090325) - The certificate chain was issued by an authority that is not trusted.>

***social.ebusinessworkshop.co.uk***
<simpleError in curl::curl_fetch_memory(url, handle = handle): Could not resolve host: social.ebusinessworkshop.co.uk>

***toot.theresnotime.io***
<simpleError in curl::curl_fetch_memory(url, handle = handle): schannel: next InitializeSecurityContext failed: SEC_E_ILLEGAL_MESSAGE (0x80090326) - This error usually occurs when a fatal SSL/TLS alert is received (e.g. handshake failed). More detail may be available in the Windows System event log.>

server_count

Hmmm something went wrong… Quick fix:

server_count$server_user_n2 <-
  server_count$server_user_n |> sapply(\(x) ifelse(length(x) == 1, x[[1]], NA))

server_count

A couple of histograms:

server_count |>
  select(n, server_user_n2) |>
  na.omit()  |>
  pivot_longer(cols = everything(),
               names_to  = "key",
               values_to = "value") |>
  mutate(nice_name = case_when(key == "n" ~ "Followers",
                               key == "server_user_n2" ~ "Users on server")) |>
  ggplot(aes(value)) +
    facet_wrap(~ nice_name, scales = "free") +
    geom_histogram(bins = 40) +
    labs(x = "Users", y = "Freq")

A scatterplot:

server_count |>
  na.omit() |>
  mutate(home = ifelse(server == "tech.lgbt",
                                 "My home server",
                                 "Elsewhere"),
         home = factor(home,
                       c("My home server", "Elsewhere"))) |>
  ggplot(aes(y = log(n, 10),
             x = log(server_user_n2, 10),
             colour = home)) +
  geom_point() +
  scale_colour_manual(values = c("magenta", "black")) +
  #theme_bw() +
  labs(y = expression(log[10]~(followers)),
       x = expression(log[10]~(total~server~users)),
       title = "Follower count by server",
       colour = "")

There is indeed a correlation:

cor.test(~ n + server_user_n2, data = server_count, method = "kendall")

    Kendall's rank correlation tau

data:  n and server_user_n2
z = 6.0364, p-value = 1.576e-09
alternative hypothesis: true tau is not equal to 0
sample estimates:
      tau 
0.3970694 

Last run (or at least knitted) Mon Aug 7 21:42:09 2023.