Synthetic_Data_Evaluation.Rmd

---
title: "R Notebook"
output: html_notebook
---
```{r}
source("../scripts/Equity_metrics.R")
source("../scripts/sunburst_process.R")
source("../scripts/table_process.R")
library(plotly)
library(formattable)
```


Read in the datasets
```{r}
ATUSreference<-read.csv(file = "../data/Atus/atus_train.csv")
ATUSsynthetic<-read.csv(file = "../data/Atus/atus_train_synthetic.csv")

MIMICRacereference<-read.csv(file = "../data/Mimic/mimic_3.csv")
MIMICRacesynthetic<-read.csv(file = "../data/Mimic/mimic_3_synthetic.csv")
```


Preprocess the ATUS data
```{r}
ATUSreference_processed<-ATUSreference %>% 
  group_by(TESEX, TEAGE) %>% 
  summarise(background_n = n())
colnames(ATUSreference_processed)<-c("Gender","Age","background_n")

ATUSsynthetic_processed<-ATUSsynthetic %>% 
  group_by(TESEX, TEAGE) %>% 
  summarise(background_n = n())
colnames(ATUSsynthetic_processed)<-c("Gender","Age","user_n")
  

ATUSreference_processed<-ATUSreference_processed %>% mutate(Gender=recode(Gender, 
                         `1`="Male",
                         `2`="Female"),
                         Age=recode(Age, 
                         `0`="15-24",
                         `1`="25-34",
                         `2`="35-44",
                         `3`="45-54",
                         `4`="55-64",
                         `5`="65-74",
                         `6`="75+"))

ATUSsynthetic_processed<-ATUSsynthetic_processed %>% mutate(Gender=recode(Gender, 
                         `1`="Male",
                         `2`="Female"),
                         Age=recode(Age, 
                         `0`="15-24",
                         `1`="25-34",
                         `2`="35-44",
                         `3`="45-54",
                         `4`="55-64",
                         `5`="65-74",
                         `6`="75+"))

```

Preprocess the MIMIC data
```{r}
MIMICRacereference_processed<-MIMICRacereference %>%
  group_by(GENDER,AGE,ETHNICITY,mortality_withinthirtydays) %>%
  summarise(background_n = n())
colnames(MIMICRacereference_processed)<-c("Gender","Age","Ethnicity","Mortality","background_n")

MIMICRacesynthetic_processed<-MIMICRacesynthetic %>%
  group_by(GENDER,AGE, ETHNICITY, mortality_withinthirtydays) %>%
  summarise(background_n = n())
colnames(MIMICRacesynthetic_processed)<-c("Gender","Age","Ethnicity","Mortality","user_n")


MIMICRacereference_processed$Mortality<-as.factor(MIMICRacereference_processed$Mortality)
MIMICRacesynthetic_processed$Mortality<-as.factor(MIMICRacesynthetic_processed$Mortality)


MIMICRacereference_processed<-MIMICRacereference_processed %>% mutate(Gender=recode(Gender, 
                         `M`="Male",
                         `F`="Female"),
            Mortality =recode(Mortality, 
                         `0`="Alive",
                         `1`="Died"))

MIMICRacesynthetic_processed<-MIMICRacesynthetic_processed %>% mutate(Gender=recode(Gender, 
                         `M`="Male",
                         `F`="Female"),
            Mortality =recode(Mortality, 
                         `0`="Alive",
                         `1`="Died"))
```


Evaluation on ATUS data 
```{r}
test_sunburst2(ATUSreference_processed,c("Gender","Age"),ATUSsynthetic_processed, "ATUSEquity1.csv", sig_t = 0.05,lower_t = -log(0.9), upper_t = -log(0.8) )
```

Evaluation on MIMIC data
```{r}
test_sunburst2(MIMICRacereference_processed,c("Mortality","Ethnicity","Age","Gender"),MIMICRacesynthetic_processed,"MIMICRaceEquity1.csv", sig_t = 0.05,lower_t = -log(0.9), upper_t = -log(0.8))
```

Tables for ATUS and MIMIC with significance threshold =0.05, lower metric threshold = -log(0.9), upper metric threshold = -log(0.8), metric = log disparity
```{r}
generate_table(0.05,-log(0.9),-log(0.8), "LDI", "ATUS")
generate_table(0.05,-log(0.9),-log(0.8), "LDI", "MIMIC")
```