sentence-by-race-pdf.Rmd

---
header-includes:
- \input{preamble.tex}
fontsize: 10pt
output: pdf_document
sansfont: RobotoCondensed
font:     RobotoCondensed
geometry: "left=1in,right=1in,top=0.35in,bottom=0.6in"
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = FALSE,
                      warning = FALSE,
                      message = FALSE,
                      dev = "cairo_pdf")

library(ojodb)

last_quarter_end <- (floor_date(Sys.Date(), "quarter") - days(1)) %>% 
  as.character

connect_ojo()
sd <- ojo_tbl("doc_sentences") %>% 
  filter(js_date >= "2009-07-01",
         js_date <= last_quarter_end,
         !is.na(doc_incarcerated_term_yrs)) %>%
  left_join(ojo_tbl("doc_offense")) %>% 
  collect()

pd <- ojo_tbl("doc_profile") %>% 
  filter(doc_num %in% !!sd$doc_num) %>% 
  collect()

disconnect_ojo()

race_pop <- read_csv("County populations by race, 2010-2018.csv")

sents <- sd %>% 
  mutate(doc_num = as.integer(doc_num),
         fy = date_to_fy(js_date),
         county = str_remove(doc_sentencing_county, " COUNTY.*") %>% 
           str_trim) %>% 
  group_by(doc_num, county, fy) %>%
  arrange(desc(doc_incarcerated_term_yrs)) %>% 
  slice(1) %>% 
  left_join(pd %>% 
              mutate(doc_num = as.integer(doc_num)) %>% 
              select(doc_num, 
                     sex,
                     race))

```

\ojologo{}
\ojotitle{Oklahoma DOC Equity Report}

# Male and female sentences
## Sentences by race and fiscal year, all counties

```{r}
sent_table <- function(county_filter = unique(sents$county), sex_filter = c("M", "F")) {
  t <- sents %>%
    filter(county %in% county_filter, sex %in% sex_filter) %>% 
    ungroup %>% 
    count(year = fy, race) %>%
    pivot_wider(names_from = year, values_from = n) %>% 
    arrange(race)
  
  return(t)
}
  

perc_table <- function(d) {
  t <- {{d}} %>%
    pivot_longer(cols = matches("\\d"), values_to = "n") %>% 
    group_by(year = name) %>% 
    summarize(`Percent White` = sum(n[which(race == "WHITE")])/sum(n, na.rm = TRUE),
              `Percent Black` = sum(n[which(race == "BLACK")])/sum(n, na.rm = TRUE)) %>% 
    pivot_longer(matches("perc")) %>%
    mutate(value = paste0(round(value*100, 1), "%")) %>% 
    pivot_wider(names_from = year, values_from = value)
  
  return(t)
}

amf <- sent_table()
knitr::kable(amf)
knitr::kable(perc_table(amf))
```

## Sentences by race and fiscal year, Tulsa County

```{r}
tmf <- sent_table("TULSA")
knitr::kable(tmf)
knitr::kable(perc_table(tmf))
```

## Sentences by race and fiscal year, Oklahoma County

```{r}
omf <- sent_table("OKLAHOMA")
knitr::kable(omf)
knitr::kable(perc_table(omf))
```

\pagebreak

# Female sentences
## Female sentences by race and fiscal year, all counties

```{r}
af <- sent_table(sex_filter = "F")
knitr::kable(af)
knitr::kable(perc_table(af))
```

## Female sentences by race and fiscal year, Tulsa County

```{r}
tf <- sent_table("TULSA", sex_filter = "F")
knitr::kable(tf)
knitr::kable(perc_table(tf))
```

## Female sentences by race and fiscal year, Oklahoma County

```{r}
of <- sent_table("OKLAHOMA", sex_filter = "F")
knitr::kable(of)
knitr::kable(perc_table(of))
```

\pagebreak

# Male sentences
## Male sentences by race and fiscal year, all counties

```{r}
am <- sent_table(sex_filter = "M")
knitr::kable(am)
knitr::kable(perc_table(am))
```

## Male sentences by race and fiscal year, Tulsa County

```{r}
tm <- sent_table("TULSA", sex_filter = "M")
knitr::kable(tm)
knitr::kable(perc_table(tm))
```

## Male sentences by race and fiscal year, Oklahoma County

```{r}
om <- sent_table("OKLAHOMA", sex_filter = "M")
knitr::kable(om)
knitr::kable(perc_table(om))
```