DSCI_522_Group_34

Contributors: Kangbo Lu, Siqi Zhou, Mitchie Zhao, Mengyuan Zhu

A data analysis project of group 34 for DSCI 522 (Data Science workflows); a course in the Master of Data Science program at the University of British Columbia.

About

Here we attempt to conduct a two-tailed permutation test to answer a statistical research question, that is, whether the number of graffiti per location of Vancouver's downtown area differs from the number of graffiti per location of the Vancouver Strathcona area. We proposed to sequentially do exploratory data analysis, determine what features and columns to be retrieved to support our permutation testing, and attach with a suitable test flavour as median to verify whether the median number of graffiti per location of Vancouver's downtown area differs from Vancouver's Strathcona area. After conducting exploratory data analysis and hypothesis testing, the results show there is no statistically significant difference between the median of counts of graffiti per recorded location in these 2 areas in Vancouver since the p-value is 1 and it's larger than the significance level of 0.05.

In the research project, the dataset provides information on the location of sites with graffiti as identified by the Vancouver city staff. The graffiti location data is sourced from the Vancouver Open Data Portal and it can be found here, specifically this file. As for the data schema, there are three columns related to our research interest. The columns are named as "COUNT", "GEO LOCAL AREA" and "GEOM". We utilized the "COUNT" and the "GEO LOCAL AREA" columns to conduct a permutation test with the difference in medians to study the graffiti situation in the Vancouver Downtown area and the Vancouver Strathcona area.

Report

The final report can be found here.

Project Collaboration

We created the following 4 files that are important for collaboration:

Usage

There are two suggested ways to run this analysis:

1. Using Docker

note - the instructions in this section also depends on running this in a unix shell (e.g., terminal or Git Bash)

To replicate the analysis, install Docker. Then clone this GitHub repository and run the following command at the command line/terminal from the root directory of this project:

docker run -it --rm -v /$(pwd):/home/project kbludocker/vancouver-graffiti make -C home/project all

To reset the repo to a clean state, with no intermediate or results files, run the following command at the command line/terminal from the root directory of this project:

docker run -it --rm -v /$(pwd):/home/project kbludocker/vancouver-graffiti make -C home/project clean

2. Without using Docker

To replicate the analysis, clone this GitHub repository, install the dependencies listed below, and run the following commands at the command line/terminal from the root directory of this project:

make all

To reset the repo to a clean state, with no intermediate or results files, run the following command at the command line/terminal from the root directory of this project:

make clean

Makefile Dependency Diagram

Click the image below to view it with the original size.

Dependencies

Python 3.8.3 and Python packages:
- docopt==0.6.2
- requests==2.23.0
- pandas==1.1.1
R version 4.0.2 and R packages:
- knitr==1.29
- docopt==0.7.1
- tidyverse==1.3.0
- ggplot2==3.3.2
- RCurl==1.98.1.2
- infer==0.5.3
GNU make 3.81

License

References

Modern Dive: An Introduction to Statistical and Data Sciences via R by Chester Ismay and Albert Y. Kim. https://moderndive.com/index.html.

Quantile estimation by Thomas Bzik. https://www.astm.org/SNEWS/images/ja14_dp.pdf.

The missing question in supervised learning blog post by Vincenzo Coia. https://github.com/vincenzocoia/vincenzocoia.github.io/blob/b9a5b5023e7d3dc0b135c032c7b1aa67376240db/_posts/2018-02-18-mean.md.

“Graffiti.” City of Vancouver Open Data Portal, 3 Feb. 2020, https://opendata.vancouver.ca/explore/dataset/graffiti/information/.

KangboLu/Vancouver-Graffiti-Analysis