---
title: "parameters_thresholds_comparisons"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{parameters_thresholds_comparisons}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)
```

## Parameters comparisons using bootstrap 

You need to specify that would like to perform `paircomparisons` (in older versions of `quickpsy` the comparisons are performed by default)

```{r fig.height=2, fig.width=6}
library(quickpsy)

fit <- quickpsy(qpdat, phase, resp, grouping = c("participant", "cond"),
                paircomparisons = TRUE, 
                B = 30) 

plot(fit, color = cond)
```

All the possible pair comparisons across parameters are include in the output data frame `par_dif`. If, for example, we want to compare the parameters across conditions for each participant we need the following subset of the data frame 
```{r message=FALSE}
library(dplyr)

fit$par_dif %>% filter(participant == participant2)
```

`dif` corresponds to the difference in parameters. `difinf`and `difsup` correspond to the limits of the bootstrap confidence intervals (default 95%). Confidence intervals that do not include the zero suggests than the parameters differ. In this case `p1` (the parameter of location) seems to differs across conditions for all participants. `p2` only seems to differ across conditions for Participant 1. 

To reduce computation time, we only included 30 bootstrap samples, but you will need to include much more. 

## Threholds comparisons using bootstrap 

The bootstrap differences across thresholds are included in `thresholds_dif`

```{r}
fit$thresholds_dif %>% filter(participant == participant2)
```