代做GGR 376、代写Java/C++编程
Assignment 2: Spatial Autocorrelation and Regression
Due Date: February 28th, 2025
GGR 376
Dataset Summaries:
Dataset Format Description
Transit Shapefiles .shp Different kinds of transit data – rail lines, rail stops, bus stops. You will need to aggregate by region. One suggestion is to use rail intersections.
Region Shapefiles .shp Shapefiles by region (can be used for thefts, complaints, and transit stations)
Land Use Shapefiles .shp Land use types, needs translation
Population by Age .xlsx In small area measurements
Daytime Pop .xlsx Daytime population by region – might reflect where commuters go.
Boarding and Off Subway .csv Rider counts for each subway stop, you will need to add this to the transit shapefile (I believe there is a rail stop file).
Pollution Complaints 2016 .xlsx Number of complaints by city by complaint type. This one is really cool.
Bicycle Thefts .xlsx Each row is one theft. Data is partially translated for convenience.
Hint: Summarize by city first. (use the numeric id)
Network .csv Node and Link CSVs, very similar to tutorial, however there is a geography component (lat/long).
Overall Directions:
Submit a complete report, with all the answers bolded with a parenthesis after for the question and section number. For example, I collected data from Tokyo Open Data (Section 1 Question 1). The assignment should be single space, and should include the following sections: introduction, methods (explicitly describe your steps), results, discussion, and conclusion. The sections do not need to be long, but they do need to be descriptive, grammatically correct, and written in a professional tone. Any figures or tables should be placed in the body of your report, any code can be placed in an appendix at the end of your report. The data visualizations in this report are going to be reviewed by peers in class on March 6, so be sure to do your best work, and make your visualizations unique.
In-text citations and a reference list are required. The format of the references should be either Nature or APA.
Section 1: Data Preparation (1 pt)
Clean datasets to be the same geography and then create some exploratory data visualizations. Create two exploratory visualizations (e.g., transit maps, land use maps, etc.).
Suggestions: Count number of crimes in Tokyo districts, complaints in districts, map transit, merge transit stops with transit ridership data.
Section 2: Spatial Autocorrelation (1 pt)
Using lecture 5 code, examine the spatial autocorrelation of 1-2 of the variables you have selected for your analysis. For example, bicycle thefts aggregated by city. Are theft rates related to geography in Tokyo?
To conduct this test, use Moran’s I. If you want to test local autocorrelation, you are welcome to use local Moran’s I, it will strengthen the quality of your report, but it is not required.
Don’t forget, this data must be spatial. Yes, it CAN be points, but you will most likely need to merge your dataset onto a spatial dataset for this analysis.
Section 3: Spatial Modelling (1.5 pts)
Develop a preliminary spatial model (spatial autoregressive or spatial error) that examines the relationship between two selected variables. Code will be provided on Quercus for these models.
Hints: They must be at the same geography.
Section 4: Data Visualization (1.5 pts)
Create at least two formal visualizations to tell your overall hypothesis. For example, my hypothesis could be that city bike thefts are predicted by daytime population counts. I will also control for the number of subway lines, and the number of people over 65, as I am assuming these individuals are not working.
Section 5: Final Report (5 pts) (0.5 for each section, listed in the overall directions)
2.5 points for quality of writing and synthesis of data sources to convey a point. (3-point deduction for incomplete/ implausible references). A lack of references is an academic offense. If you use translation tools, please reference them. Including ChatGPT.
Bonus Section: Network Analysis (Max: 2 pts)
Using the network dataset and the mini tutorial on social networks, create a network visualization and relate it to your report. If you choose to do this portion of the assignment, then include it as part of your analysis in your report.