Assignment 2: Exploratory Data Analysis

Note: before plotting, the data was slightly cleaned up to get rid of some extranneous coordinates outside of the city limits.

An easy answer to our first question is quite simple: crimes occur everywhere. Plotting the coordinates of every reported crime since 2001 provides a clear approximation of a map of the city of Chicago. A quick check of an actual map shows that the blank spots largely consist of airport tarmacs, bodies of water, and interstates.

Before delving more deeply into the "where" of crime in Chicago, let's look at a breakdown of the different types of crime that have occurred.

The standouts here are theft, battery, criminal damage, and narcotics. Our guess would be that many of the criminal damage incidents occur in tandem with others. For this reason, for the remainder of the analysis the crimes will be grouped into four sets:

Violent crimes other than homicide (assault, battery, and robbery)
Narcotics
Theft and burglary
Homicide

Now, let's go back to our previous map and break it down by the type of offense.

This gives quite a powerful insight into the different types of crime that occur in different parts of Chicago. In particular, by comparing the maps we can see some geographic correlation between different types of crime. For instance, it appears that everywhere where there is a strong prevalence of violent crime, homicides are more common there than in other places. Incidents involving narcotics tend to only occur in areas with a high volume of violence, but the converse is not necessarily true. Theft and narcotics do not appear to have much correlation.

Notice the diagonal swath of land on the north-eastern side of the city (most clearly seen by the absence of homicides). From this series of maps, it appears to be one of the areas with the lowest incidence of crime, aside from some theft near the lake. What is special about this area?

By and large, this area most devoid of homicides and violent crime over the past 17 years is correlated extremely highly with median household income. Perhaps unsurprisingly, there is a relatively large volume of theft in this area.

Now let's move on to our second question: when do crimes occur?

Here are periodic weekly cycles for each type of crime. Notice that most types of crime have a roughly equivalent period in each day. For narcotics offenses, there is a peak around noon and a larger/steeper peak closer to midnight. In constract, violent offenses build from the early morning throughout the day in more or less one single peak until tapering off near midnight (with the exception of Saturday nights). Thefts largely occur at noon more than any other time. The outlier to daily periodicity within the weak being roughly constant is in homicides; most frequently, they occur specifically after midnight on Friday and Saturday nights.

Next, we turn to how crime has changed over time, our fourth question. From here onward, we will ommitt offenses from 2018 so as not to skew results. To begin, consider the following bar chart.

It is clear that crime has seen a considerable decrease since 2001. Let's break this down by the type of offense once more, first looking at homicides.

Although the past few winters are concerning and May of 2001 might have had an unusually low number of homicides, we see here that year-over-year there has been a large, steady reduction in the number of homicides reported.

Since homicides happen relatively rarely, a month-by-month basis appeared to be most appropriate for the above graphic. For narcotics and violent crimes, their large numbers lend themselves to a day-by-day analysis.

The comparison to the previous graphic is somewhat staggering. Although there has been a large decrease over the last 17 years in narcotics offenses, there was a massive uptick in the mid-2000's. Curiously, there seems to be a larger swing to this trend occuring in late summer/early fall.

Next, let's look at violent offenses.

The same trend is apparent here, though not to such a large degree and for a considerably shortly duration. By and large, violent offenses showed a mild increase and then a steady decrease over the last 17 years.

It appears that for many parts of Chicago, when considering whole half-decades at a time, many areas actually saw a large increase in total crime in the latter half of the 2000's, but have more recently seen improvements.

Finally, we'll tie our main two approaches together by bringing the type of offense into the picture.

As suggested independently by our previous analyses, narcotics crimes truly dominated in terms of growth in the early 2000's. Luckily, that sector has also made the largest reduction in the time since. Unforunately, the homicides of the last few winters mentioned above also appear to have left their quite a mark.

Summary

How has Chicago fared for the past 17 years? We see the suggestion of either a drug epidemic or a pronounced crack down on drug use roughly 15 years ago, but there has been a large decrease in overall offenses. The expected trend of some neighborhoods being more associated with crime is very prevalent here (especially in the case of violent crimes).

Finally, if you are in Chicago and want to avoid crime as much as possible, conduct the majority of your business around 4am.