Static Analysis

Last Updated (8/26)

This repository must be placed within the "automated-assessments" repository so that the automated graders fetched for analyzing.

Additionally, python3, pip3, and the supporting modules must be installed. The required modules can be found in requirements.txt. All modules in this file can be installed recursively by running pip3 install -r requirements.txt.

If additional modules are required later on, it is recommended to set up a virtual environment, installing virtualenv, so that requirements.txt can be updated with only modules specific to this project.

run.py: python3 run.py project-name [optional keyword] [--verbose/-v]

• Requires a project name, i.e. encore, act-one
• Second optional argument to narrow down to specific modules within a project, i.e. animation-L1. If no argument is given, run.py will run on all modules in a project
• For each row (classroom) in the CSV metadata file, reads the Scratch studio URL for each module which the analysis is being conducted on
• Uses webScrape.py to scrape the studio jsons if they are not already present locally
• Grades each project using grade.js
• Runs dataProcessing.py and plot.py to output csvs and graphs
• Creates any directories required but a file called "metadata.csv" should be made before running

dataProcessing.py (ran by run.py): python3 dataProcessing.py (module)

• Manipulates the data outputted by grade.js and outputs aggregated graphs

plot.py (ran by run.py): python3 plot.py (module)

• Given the CSVs from dataProcessing.py, creates the following graphs:
  • Per grade:
    • Distribution, classroom completion per requirement, and classroom completion totals
  • Overall:
    • Classroom completion by grade and teacher/classroom
• Graphs are grouped as PDFs in the graphs folder and saved individually as PNGs within graphs/pngs

webScrape.py (helper for run.py):

• Scrapes the web for a scratch project, adding it to the folder json_files_by_studio and within the folder of its corresponding Studio
• Files are named by the Scratch username of the user. If they the project has been remixed by a Scratch Encore account, it uses the username of the project that it was remixed from

grade.js (helper for run.py):

• Grades each json in a Studio directory (within json_files_by_studio)
• Adds results from each project as rows of a CSV
• CSVs, for each classroom (Studio), are placed in folders by their corresponding grade

Only the files mentioned above have been used for static analysis for Scratch Encore during Summer 2019