Final.Rmd

---
title: "Final Project"
output:
  pdf_document: default
  html_document: default
---

```{r setup, include=FALSE}
library(usmap)
library(ggplot2)

table12 <- read.csv(file="table12.csv")
table11 <- read.csv(file="table11.csv")
```

## R Markdown


```{r}
summary(table12)

table12$Number.of.participating.agencies <- 
  as.integer(gsub(",", "", table12$Number.of.participating.agencies))
table12$Population.covered <- 
  as.integer(gsub(",", "", table12$Population.covered))
table12$Total.number.of.incidents.reported <-
  as.integer(gsub(",","",table12$Total.number.of.incidents.reported))

summary(table12)
```

<First check if there is a correlation with the number of agencies participating
and the population>
```{r pressure, echo=FALSE}
options(scipen=999)
plot(table12$Population.covered, table12$Number.of.participating.agencies, 
     xlab="Population", ylab="Number of Agencies")
plot(table12$Population.covered, table12$Number.of.participating.agencies, 
     xlab="Population", ylab="Number of Agencies",
     xlim = range(0, 12000000)) 

cor(table12$Population.covered, table12$Number.of.participating.agencies)
```

<See how many of agencies are submitting reports out of number submitting>
<data shows that a very low number agencies are actually submitting reports>
```{r}

percent_participate_table <- data.frame(table12$Participating.state.Federal,
                                  table12$Agencies.submitting.incident.reports,
                                  table12$Number.of.participating.agencies)

colnames(percent_participate_table) <- c('States', 'Agencies Submitting Reports'
                                         , "Agencies Participating")

percent_participate_table = transform(percent_participate_table, 
                            freq = (percent_participate_table$`Agencies Submitting Reports`/percent_participate_table$`Agencies Participating`)*100)

```



<there is a correlation -> Lets see if we can predict the n>
<first lets see if the number of ag>
```{r}
plot(table12$Number.of.participating.agencies, table12$Total.number.of.incidents.reported, 
                                  xlab="Population", ylab="Number of Agencies")
plot(table12$Number.of.participating.agencies, table12$Total.number.of.incidents.reported, 
                                  xlab="Number of agencies", 
                                  ylab="Number of incidents", 
                                  xlim = range(0, 800))

cor(table12$Number.of.participating.agencies, table12$Total.number.of.incidents.reported)
```
<lets see if there is an increase of agencies means an increase of reports>
```{r}
partagent <-table12$Number.of.participating.agencies
reported <-table12$Total.number.of.incidents.reported
lm_agencies_report <- lm(formula = partagent ~ reported,
                         data = table12)
summary(lm_agencies_report)
```

<linear regression assumptions test>
```{r}

plot(partagent, lm_agencies_report$residuals, main="Residuals vs. x", xlab="x, Scaled speed", ylab="Residuals")
abline(h = 0, lty="dashed")
```

```{r}
plot(lm_agencies_report, which=3)
```

```{r}
plot(lm_agencies_report, which=5)
plot(lm_agencies_report, which=2)
```
<a regression isn't valid so we can't say that for every predicted agencies that we get more reports>

<now look at the number of offenses we are seeing across state lines>
```{r}
summary(table11)
```
<total offences in general looking at freq from table 12 to table 11>
```{r}
table11_total = data.frame(table11$Participating.state.Federal, table11$Total.offenses, percent_participate_table$freq)
```

<most common crimes per state>
```{r}
table11_maxperstate <- data.frame(table11$Participating.state.Federal)
table11_maxperstate['max'] <- apply(table11[3:15], 1, max)
table11_maxperstate['highest occuring crimes'] <- colnames(table11[3:15])[max.col(table11[3:15], ties.method = "first")]

table11_total
table11_maxperstate
```

```{r}
incidents <- data.frame(table12$Participating.state.Federal, as.numeric(gsub(",","",table12$Total.number.of.incidents.reported)))
colnames(incidents) <- c('state', 'Total number of incidents')
plot_usmap(data = incidents, values = "Total number of incidents")+scale_fill_continuous(name = "Number of Incidents",low = "white", high ="darkblue", label = scales::comma) + theme(legend.position = "right")
```
```{r}
percent_participate_table <- data.frame(table12$Participating.state.Federal,
                                  table12$Agencies.submitting.incident.reports,
                                  table12$Number.of.participating.agencies)

colnames(percent_participate_table) <- c('state', 'Agencies Submitting Reports'
                                         , "Agencies Participating")

percent_participate_table = transform(percent_participate_table, freq = (percent_participate_table$`Agencies Submitting Reports`/percent_participate_table$`Agencies Participating`)*100)


plot_usmap(data = percent_participate_table, values = "freq")+scale_fill_continuous(name = "Percent of Participating Agencies",low = "white", high ="darkblue", label = scales::comma) + theme(legend.position = "right")
```

```{r}

colnames(table11_maxperstate) <- c('state', "max", "incidents")
plot_usmap(data = table11_maxperstate, values = "incidents") + theme(legend.position = "right") + scale_fill_brewer(name = "Incidents with the highest occurrence",type = 'qual', palette = 1) 
```