---
title: Prediccion a largo plazo de las variables de verbos
output: html_document
---

```{r warning = FALSE}
date() 
```

Lo que hemos hecho es seleccionar algunas variables (NP) que indican la fluencia verbal y ver si hay algun cambio significativo en el diagnostico en la visita 5

### Plots

```{r message = FALSE, warning = FALSE}
library(ggplot2) 
fc <- read.csv("facehbi_response.csv") 
langs <- c("fe_verb_fluency_np", "kissing_dancing_images_fac", "kissing_dancing_words_fac", "action_naming_free_fac") 
for (lang in langs) {         
	out_fig <- paste0("response_",lang, ".png")         
	ggplot(fc, aes(x = syndromic_diagnosis, y = .data[[lang]], fill = syndromic_diagnosis)) + stat_boxplot(geom = "errorbar", width = 0.25) + geom_boxplot() -> p         
	print(p)
}
```

Ya que parece que hay diferencias significativas, vamos a ver que dice la estadistica,

### Stats

```{r message = FALSE, warning = FALSE}
fc <- read.csv("facehbi_response.csv") 
langs <- c("fe_verb_fluency_np", "kissing_dancing_images_fac", "kissing_dancing_words_fac", "action_naming_free_fac") 
for (lang in langs) {
  a <- t.test(as.formula(paste(lang, '~ factor(syndromic_diagnosis)')), data = fc, na.action = "na.omit")
  print(lang) 
  print(a)
}
```