jonspring

# Inspired by Nadieh Bremer's lasso annotations here: https://twitter.com/NadiehBremer/status/1277622602735865856
# h/t Cedric Scherer: https://twitter.com/CedScherer/status/1278351840074240001


library(ggplot2); library(ggforce)

add_lasso_layer <- function(x1, x2, y) {
  width = x2 - x1
  space = width * 0.05
  lasso_points = data.frame(

jonspring

devtools::install_github("coolbutuseless/poissoned")
library(poissoned)

# v1
poisson_disc(ncols = 100, nrows = 50, cell_size = 10) %>%
  as_tibble() %>%
  arrange(y) %>%
  mutate(rando = runif(n()),
         visibility = case_when(
           y >= 300   ~   (500-y)/200,

jonspring

library(tidyverse)

example_df <- tribble(
  ~Year, ~Category, ~Name,
  2015,  "A",       "1",
  2015,  "A",       "3",
  2015,  "A",       "5",
  2015,  "C",       "2",
  2015,  "C",       "4",
  2016,  "A",       "1",

jonspring

library(palmerpenguins)
library(tidyverse)
library(umap)
library(gganimate)

penguins_clean <- penguins %>% filter(!is.na(bill_length_mm), !is.na(sex))
penguinos <- function(neigh = 13) {
  penguins_clean %>%
    select(bill_length_mm:body_mass_g) %>%
    scale() %>%

jonspring

library(tidyverse)
library(fable)
library(gganimate)
set.seed(1)
df <- data.frame(x = 1:365,
                 y = cumsum(runif(365, min = -1)))

smoothy <- function(alpha1) {
  df %>%
    as_tsibble(index = x) %>%

jonspring

dog_moves <- readr::read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2019/2019-12-17/dog_moves.csv')
dog_travel <- readr::read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2019/2019-12-17/dog_travel.csv')
dog_descriptions <- readr::read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2019/2019-12-17/dog_descriptions.csv')

library(tidyverse)
library(tidytext)
library(lubridate)

doggo_names <- dog_descriptions %>%
  select(name, breed_primary, size, sex, size, contact_state, contact_zip, posted)

jonspring

# Some ideas to try out on the data from:
# https://gist.github.com/brooke-watson/ccf3d1b1f4449ab55a72f7835a52e599



# 0. Let's describe what type of data in each row
sw1_annotated <- starwars_garbage_data1 %>%
  # Counting each new group of data
  mutate(group = cumsum(v1 == "Character Name")) %>%
  # Assign rows within each group

jonspring

# https://twitter.com/jonlovett/status/1191490424965218304
# What I'm curious about is how we're distributed inside of our time zones, because the further east you are inside of a time zone, the harder standard time hits. Boston's sunset is 4:34pm today. Brutal. Detroit's sunset, on the other side of the same time zone, is 5:22pm.


# I'm a big fan of Jon Lovett's podcasts, and I love trying new things in R, so it was Game On when I saw the tweet above. I googled around and combined a few data sources to get US counties' populations, coordinates, sunsets, and time zones. From those you can see who has it worst in the evenings when standard time kicks in. 

# Curiously, big cities in the US tend to be on the east, "darker evening" end of their time zones.


####### Libraries #####

jonspring

library(rtweet)
library(tidyverse)
library(tidytext)
library(ggforce)

sw <- search_tweets("Rogue, IV", n = 90000, include_rts = FALSE)

sw_clean <- sw %>%
  unnest_tokens(word, text, drop = F, to_lower = T) %>%
  filter(word %in% c("i", "ii", "iii", "iv", "v", "vi",