mcfrank

library(tidyverse)
library(wordbankr)

eng_wg_inst <- wordbankr::get_instrument_data(form = "WG", language = "English (American)", 
                                              item_info = TRUE, administration_info = TRUE)

by_cat <- eng_wg_inst |>
  filter(item_kind == "word") |>
  group_by(data_id, lexical_category, age) |>
  summarise(prod = sum(produces), 

mcfrank

library(tidyverse)
library(pwr)


pwrs <- expand_grid(n = seq(0, 100, 10), 
            es = seq(0, 1, .1)) |>
  mutate(pwr = map2_dbl(es, n, 
                    ~pwr.t.test(d = .x, n = .y)$power))

ggplot(pwrs, aes(x = n, y = pwr, col = as.factor(es))) + 

mcfrank

library(xkcd)
library(tidyverse)

d <- tibble(age = 1:20) |>
  mutate(upper = 1e6 * age * 12 + ifelse(age > 5, 2.5e5 * 52 * (age - 5), 0), 
         lower = 1e5 * age * 12) |>
  pivot_longer(upper:lower, names_to = "bound", values_to = "vocabulary")

pdf("~/Projects/AI commentaries/scale.pdf", width = 5, height = 4)
ggplot(d, aes(x = age, y = vocabulary, col = bound)) + 

mcfrank

library(wordbankr)
library(langcog)
library(tidyverse)
library(brms)
library(forcats)
library(survey)
library(gamlss)
theme_set(theme_mikabr())
font <- theme_mikabr()$text$family

mcfrank

# relies on zoo package

# function to find the baseline value
baseline_looking <- function (lts) {
  # select lts > 12s
  lts_12 <- lts[lts > 12 & !is.na(lts)]
 
  # sum the first three
  baseline <- sum(lts_12[1:3], na.rm=TRUE)
 

mcfrank

library(tidyverse)
library(assertthat)

shuffle_data <- function(d) {
  d$participant_id <- shuffle(d$participant_id)
  
  return(d)
}

get_knower_level <- function (q, r, technique = "perfect") {

mcfrank

# approach from
# https://www.medrxiv.org/content/10.1101/2022.11.02.22281762v1.full.pdf

library(tidyverse)
library(survival)
library(ggsurvfit)
#devtools::install_github("maxlinde/baymedr")
library(baymedr)
# library(BayesSurvival)

mcfrank

---
title: "mtcars example markdown"
author: "Mike Frank"
date: "2023-03-21"
output: 
  html_document:
    toc: true
    toc_float: true
---

mcfrank

library(tidyverse)
library(lme4)

sgf <- read_csv("https://raw.githubusercontent.com/langcog/experimentology/main/data/tidyverse/stiller_scales_data.csv") |>
  mutate(age_group = cut(age, 2:5, include.lowest = TRUE), 
         condition_f = factor(ifelse(condition == "Label", 
                                     "Experimental", "Control")),
         age_centered = age - mean(age))

mod1 <- glmer(correct ~ age * condition + (1|subid) + (1|item), 

mcfrank

# starts at line 716 of paper.Rmd

d_lmer_scale <- d %>%
  filter(trial_type != "train") %>%
  mutate(log_lt = log(looking_time),
         age_mo = scale(age_mo, scale = FALSE),
         trial_num = trial_num - 8.5, 
         item = paste0(stimulus_num, trial_type)) %>%
  filter(!is.na(log_lt), !is.infinite(log_lt))