---
title: "The relation between DayOfWeek and the number of crimes"
author: "Orange81"
date: "11 november 2015"
output: html_document
---

```{r}
#load the libraries
library(ggplot2)
```

```{r}
#load the data
train <- read.csv("../input/train.csv", header = T)
```

Lets start with all crimes combined
```{r}
# get number of incidents per DayOfWeek
weekday <- as.data.frame(table(train$DayOfWeek))
names(weekday) <- c("DayOfWeek", "Frequency")

# set the levels in order we want
weekday$DayOfWeek <- factor(weekday$DayOfWeek, 
                              levels=weekday$DayOfWeek[order(weekday$Frequency, 
                                                             decreasing = T)])
# make plot
plot <- ggplot(data=weekday, aes(x=DayOfWeek, y=Frequency)) 
plot + geom_bar(stat="identity") + labs(title="All Categories combined")
```
As shown in the figure above, most crimes happen on friday.


Now differentiate per crime category
```{r}
## Per Category

plots <- list()  # new empty list
for (i in 1:length(unique(train$Category))) {
        weekday <- as.data.frame(table(train$DayOfWeek[which(train$Category==unique(as.character(train$Category))[i])]))
        names(weekday) <- c("DayOfWeek", "Frequency")
        
        # set the levels in order we want
        weekday$DayOfWeek <- factor(weekday$DayOfWeek, 
                                    levels=weekday$DayOfWeek[order(weekday$Frequency, 
                                                                   decreasing = T)])
        # make plot
        plot <- ggplot(data=weekday, aes(x=DayOfWeek, y=Frequency)) + geom_bar(stat="identity") + labs(title=unique(as.character(train$Category))[i]) + ylim(0,30000)
        plots[[i]] <- plot
}

plots
```