chrishanretty

library(countrycode)
library(tidyverse)
library(rio)

holder <- list()

### (0) PartyFacts lookup
pf <- read.csv("data/partyfacts-external-parties.csv")

### (1) Manifesto Project

chrishanretty

### Let's generate some data with a step function
library(mgcv)
set.seed(1116)

nDays <- 200
day <- 1:nDays

y <- rnorm(200, 1, 0.25) + I(day > 100) * -0.5

y <- ifelse(y < 0,

chrishanretty

library(sp)
library(raster)
library(viridis)
library(rgdal)
library(rnaturalearth)


### need two things, presumably
###  - a polygon mask
###  - a population raster

chrishanretty

library(tidyverse)

### Pivot the table so we have four columns
### L/R, GAL/TAN, issue area, and value
dat <- read.csv("surridge1.csv") %>%
    pivot_longer(cols = -c(Column1, Column2),
                 names_to = "Area")

### Tidy up the factor levels and order
dat <- dat %>% 

chrishanretty

library(brms)
library(mice)
library(tictoc)

### Set a nice long number of iters so we can be sure it's not the
### overhead of spinning up cores, passing data, etc.,
nIters <- 1e5
thin <- nIters / 1000

# Load data

chrishanretty

### Start by plotting stuff
library(tidyverse)
library(hrbrthemes)

## Download data from https://docs.google.com/spreadsheets/d/1inaI17l9Q-XADRNE_toXLp65sKu4_x4bXUFjZ3qjB2o/edit?usp=sharing
dat <- read.csv("wherever_i_saved_the_google_docs.csv")

### Estimate a simple model which we'll use when graphing
mod <- glm(I(Funded == "Yes") ~ Rank2 + ConWinner1,
           family = binomial,

chrishanretty

library(tidyverse)
library(rdrobust)
library(hrbrthemes) 
library(cowplot)
library(Cairo)

dat <- readRDS("tidy_data.rds")

dat <- dat %>%
    mutate(count_satisfied.20 = round(pct_satisfied.20 * response.20),

chrishanretty

{
  "class": "Survey",
  "id": 609340,
  "internal_title": "Learning under lockdown survey",
  "l10n": "en_GB",
  "runs": [
    {
      "class": "SurveyRun",
      "alert_frequency": 0,
      "alert_last_update": 0,

chrishanretty

dat <- read.csv("https://www.dropbox.com/s/e3p6xvjq6desua8/am_i_going_crazy.csv?dl=1")
holder <- list()
for (i in 1:20) {
    holder[[i]] <- lm(lasso ~ ols, data = dat)
}

### For me, 
lapply(holder, coef)
### gives the following. Note the intercepts in runs 1, 5 and 8 are radically different

chrishanretty

library(tidyverse)
###
nGrp <- 250
grp_ideal_points <- rnorm(nGrp, 0, 1)
grp_sd <- 1
ind_sd <- 1
party_pos <- -1

joinfunc1 <- function(ind_pos, party_pos) {
    ## Probability joining party