This week in class, we added Sumbul to our group!

Originally, Lavania and I had planned to answer two questions with GIS. This week we decided to add two more questions, incorporating some statistical coding, for a total of four questions.

1. Why did some counties have protests related to labor laws over others?
2. Which organizations are involved most in violent demonstrations?
3. What is the ratio of violent protests to peaceful protests?
4. Is there a year that stands out with more violent protests than others?

 

As I will be answering the first two questions using GIS, I took class time to edit the data so that it could be mapped more effectively.

Today in class we discussed point pattern and point process and how they can be used to find randomness in our data.  I also got a chance to go over my questions in class and Gary suggested that I narrow down the scope from states to counties as they are easier to digest. I also went over my questions with Lavanya and we created this list of questions to narrow down later for our assignment:

1. Which demonstrations took place in response to gun violence and where did they occur?
2. Why did some counties have protests related to labor laws over others?
3. Which organizations are involved most in violent demonstrations?
4. which counties have the highest number of incidents?
5. are there any hotspots of violent activity?

The bolded questions peaked our interest the most.

This week in class we talked about k-clustering which is used a lot in spatial analysis. We also reviewed the new data that we will be working with for our next project. When I looked through the data, as Garry suggested I took out the peaceful protests and put them into a separate tab in the excel file to be studied later. My very first questions when looking at the data is what constitutes a protest with intervention? I was wondering if those are protests where police intervened during a peaceful protest.

 

Other questions while looking at the data are:

1. which are the organizations involved?
2. Do those organizations, organize around certain locations or is it random?
3. Which states have the highest number of incidents?
4. What are the most common event types in different regions?
5. Are there any hotspots of violent activity?

This week in class, Lavanya and I wrapped up our Project One report. We did a final read-through to correct errors and add any missing elements. Overall, the project was insightful, and our findings surprised me—I did not expect to find no correlation between the time of year and shootings of mentally ill individuals. On the other hand, I was not surprised by the lack of overlap between areas with high body-camera usage and shootings of mentally ill individuals.

Today in class we discussed Mahnoor’s project and the time series that she is working on. I found it interesting how she narrowed the data per state and then broke it down further by race. It is quiet impactful to see the data visually.

 

Regarding my project:

I have created heat maps representing:

• Locations where mentally ill individuals were shot and body cameras were used.
• Locations where body cameras were used.

(To create these maps, I used Esri’s ArcGIS Online program.)

When comparing both maps, I noticed that although body cameras were most frequently used in densely populated areas, the recorded instances involving individuals with mental illness are significantly fewer.

 

This week we did not have class, however I was able to get started working in Mathematica. My code for my 2 questions, “Where are the body cameras used more? (which county)” and “Were there any mentally ill people shot near the locations where body cameras are used?” is coming along. I am having issues getting Mathematica to read the data correctly to be able to map them out. Going to keep at it of course…

This week I was able to download and begin working with Mathematica. It has been a little rough around the edges but I will continue to work with the program to get the answers to the questions we have. I have also begun the intro to our report.

This week in class we reviewed how Mathematica can be used in a statistical setting. I started downloading the application to get myself used to Mathematica’s interface. I have had trouble installing Mathematica and am waiting for IT to email me back with solutions. During class we discussed the questions Lavanya and I decided on. Ryan was very helpful in suggesting that we find the correlation between location of police station and where the shootings of people with mental illness as well as where the body cameras were not used. We will be taking his suggestion and applying it to our research. Hopefully by the end of next week I will have the Mathematica issue resolved so I can begin my portion of the project.

This week Lavanya and I really zeroed in on our questions and narrowed down exactly what we wanted to report about. Our questions are:

• What is the rate of women killed with mental illness vs men with mental illness?
• Is there a time of the year where people are shot more than other times?
• Where are the body cameras used more? (which state)
• Were there any mentally ill people shot near the locations where body cameras are used?

We decided that we would use python to answer the first two questions and Mathamatica or ArcGIS to answer the bottom two questions.

Hello all,

This week in class we looked through a data set from the Washington Post about deathly police shootings. Upon the first initial analysis my partner Lavanya and I organized the data by gender and immediately noticed that out of the 10,389 shootings only 4.4% of those were women. In class we made other observations such as how there was missing data and that it seemed like TX and CA had the most data. We also asked ourselves how old was the youngest victims were amongst the women and male victims. This past week while reviewing the data alone I noticed that out of the 10,389 shootings only 1,792 were recorded by police body cameras. With a quick custom short I was able to notice that none of those shootings recorded via body cams were associated with women.