05-ET_pupil_prp_lmm.Rmd

---
title: "02-pupil_size_peak_lmm"
output: html_document
date: "2023-08-22"
---

```{r setup, include=FALSE}
# install all needed packages 
pacman::p_load('dplyr', 'ggdist', 'ggeffects', 'ggpubr', 'lme4', 'emmeans', 'rstatix', 'car', 'rsq', 'sjPlot', 'MASS')
library(lme4) # for the analysis
# library(tidyverse) # needed for data manipulation.
library(emmeans)
library(lmerTest)
library(MASS)
```

##  Setting the parameters
Setting the parameters. This is where the user can change the parameters if needed

```{r Housekeeping and parameters setting}
rm(list = ls())
bids_root <- getwd()

```
# Experiment 1
## Model RT2 as a function of pupil latency
```{r Loading and transform the data}
# Generate the file name:
full_path = file.path(bids_root, 'bids', 'derivatives', 'pupil_beh_correlation', 'prp', 'pupil_RT2_correlation.csv')
data_exp1 <- read.csv(full_path)
# Remove the targets:
data_exp1 = data_exp1 %>% filter(task != "target")
# Convert the task relevance to a binary variable:
data_exp1 = data_exp1 %>% filter(SOA_lock != "offset")

# Modelling:
model_exp1 <- lmer(formula = RT2 ~ 1 + pupil_latency + (pupil_latency | sub_id), 
                         data = data_exp1)
summary(model_exp1)
# Save to file:
plot(fitted(model_exp1),residuals(model_exp1))
qqnorm(residuals(model_exp1))
```
# Experiment 2
## Model RT2 as a function of pupil latency
```{r Loading and transform the data}
# Generate the file name:
full_path = file.path(bids_root, 'bids', 'derivatives', 'pupil_beh_correlation', 'introspection', 'pupil_RT2_correlation.csv')
data_exp2 <- read.csv(full_path)
# Remove the targets:
data_exp2 = data_exp2 %>% filter(task != "target")
data_exp2 = data_exp2 %>% filter(SOA_lock != "offset")

# Modelling:
model_exp2 <- lmer(formula = RT2 ~ 1 + pupil_latency + (pupil_latency | sub_id), 
                         data = data_exp2)
summary(model_exp2)
plot(fitted(model_exp2),residuals(model_exp2))
qqnorm(residuals(model_exp2))
```