Mobile Health and Activity Monitoring

File structure

/
├── docs                           # Documents
│   └── mhealth23_exercise.pdf     ### Exercise description
├── img                            # Images used in this README
├── submission                     # Notebooks and py file for submission
│   ├── group24_pathdetection.ipynb
│   └── group24_stepcount.ipynb
├── task2                          # Task2 related files
│   ├── .gitignore
│   ├── README.md                  ### Instructions for task2 submission
│   └── trace_postproc.py          ### Trace postprocessing script
├── mapzurich.png
├── example.ipynb
├── path_detection.ipynb              # Path detection script (workspace!)
├── step_count.ipynb               # Step count notebook (workspace)
└── README.md

Things to notice for collaboration

• Mind your commit message format
  • This repo adopts the Conventional Commits format
  • The commit message should be meaningful and informative (describing what you have done)
• The master branch is protected, you should not (and cannot) push to it directly
  • ALWAYS create a new branch for your work
  • Make a pull request to merge your branch to master
  • You need at least one approval from your teammates to merge your branch to master
  • After merging, delete your branch

Working on Kaggle

The following docs from Kaggle are useful. Note that Kaggle can only import a certain file instead of the whole repo from GitHub.

• Import GitHub code from GitHub to Kaggle
• Save Notebooks (and commit directly) from Kaggle to GitHub

Before you import the notebooks from GitHub, always bear in mind that the master branch is protected. You cannot push to it directly. Thus, to save and sync your work on Kaggle to GitHub, it is recommended to create a new branch first and make edits on that branch.

Note: It is not recommended to modify the notebooks in submission directory because they are used for submission. The recommended workspace notebooks are the ones located directly in the root directory. For subtask 1, use step_count.ipynb. For subtask 3, use path_detection.ipynb.

Note that Kaggle might automatically change the underscore (_) in the file name into minus(-), which results in two almost identical files appearing on GitHub.

You can also use example.ipynb as a starting point since it is the workspace provided by the TAs.

In order to add data, the Add Data button is located in the notebook settings (which is hidden in at the bottom right cornor of the page, see the screenshot below).

Special instructions for Subtask 1

After importing the notebook you wish to work on, add the "Mobile Health 2023 - Step Count" dataset and (if you want to use our self-recorded step count dataset) the "pigwalk" dataset to your Kaggle notebook.

Special instructions for Subtask 3

After importing the notebook you wish to work on, add the "Mobile Health 2023 - Path detection Submission" dataset and the "mham_task2_submission" dataset to your Kaggle notebook.

Useful links

• Course webpage
• Kaggle competitions
  • Task 1: Step count
  • Task 3: Path detection
• Resource GitLab link
• Polybox submission links
  • Task 1
  • Task 2
  • Task 3