szeigo / pythonhomicide

Python exercise about US homicides

Synopsis

This is our solutions for the python exercise about Homicide Reports US 1980-2014. The github repo contains all python files, all charts generated as images and 2 txt files.

Questions asked

• Q1:Which ethnicity is it most common for the victims and perpetrators to be?
• Q2:Which weapon is most used by men?
• Q3:Which weapon is most used by women?
• Q4:What is the age of the youngest victim and the oldest victim?
• Q5:Average age of victims?
• Q6:Male to female ratio of perpetrators?
• Q7:Top 10 states with most homicides? display it with bars (barchart) or similar
• Q8:Are younger perpetrators (age 15-25) more likely to get caught then older ones (25+)?

Solutions

1.Which ethnicity is it most common for the victims and perpetrators to be?

There is a lot of "unknowns" in both graphs, which makes it harder to make it harder to analyze.
Hispanic are more involved in crimes than non-hispanic(Both victims and perpetrators).
Less unknown ethnicity in victim's graph. Probably easier to document the ethnicity of victims than perpetrators for unsolved cases.

2.Which weapon is most used by men?

• First thing to note is the numbers between the two graphs. Men are more often killers than women.
• Women tend to use "Knife" or "Blunt objects" more often than men.
• Women use "Knife" 0.80 for every 1.0 time they use "Handgun". (1:0.8)
• Men use "Knife" 0.30 for every 1.0 time they use "Handgun". (1:0.3)
• Women use "Blunt Object" 0.33 for every 1.0 time they use "Handgun". (1:0.33)
• Men use "Blunt Object" 0.23 for every 1.0 time they use "Handgun". (1:0.23) This might be because women isn't as gang-related as men and therefore doesn't have Handgun as often.
  Their murders might be more in the heat of the moment.

3.Which weapon is most used by women?

4.What is the age of the youngest victim and the oldest victim?

We added TXT file to the project. Didn't see any reason to display with a graph.
Here is a snippet of the code getting the age of the youngest and the oldest victim and writing to a txt file.
Youngest victim was 0 years old. Oldest guy was 998 years old. Seemed like an error in the csv file or documentation.

5.Average age of victims?

We added TXT file to the project. Didn't see any reason to display with a graph.
Here is a snippet of the code getting the average age of victims and writing to a txt file.
Average age was 35.0.

6. Male to female ratio of perpetrators?

90% male to 10% female (9:1) We used masks to remove all the "unknown", which was 70% of the perpetrators.

7. Top 10 states with most homicides? display it with bars (barchart) or similar

This graph needs to compare with population numbers of the states.
e.g. "Top 10 states with most homicides per citizen"

Heres a list of the biggest states in US by population:

top 10 homicides state	Population	In Top 10 state by population	Homicides per 100	Rank in population
1. California	39,250,017	✔	393.3	1
2. Texas	27,862,596	✔	448.7	2
3. New York	19,745,289	✔	400.7	4
4. Florida	20,612,439	✔	554.6	3
5.Michigan	9,928,301	✔	348.9	10
6. Illinois	12,801,539	✔	494.8	6
7. Pennsylvania	12,802,503	✔	528.2	5
8. Georgia	10,310,371	✔	488.9	8
9. North Carolina	10,146,788	✔	497.6	9
10. Louisiana	4,681,666		238.5	25

8. Are younger perpetrators (age 15-25) more likely to get caught then older ones (25+)?

The % of people not getting caught has so little a difference. Age doesn't affect chances of getting caught statistically.

Authors

David Shmilowitz

Jonathan Szigethy