Readme

Author: Matthew Hull

Using the Apple Watch Workout Pipeline

• All watch workouts are automatically synced through the HealthFit app on iPhone
• Sources currently are Strava, Apple Activity, and Nike+
• After sync, HealthFit uploads exports each workout as a fitfile to iCloud with the extension .fit

Use make for the following steps:

• Run fitparse_process.py to grab all fitfiles and combine them with a common format into fit.csv
• Preprocess and clean aggregated data from fit.csv using preprocessing.Rmd. Using knit to pdf on the R Markdown workbook produces exploratory plots as well.
• Output of the preprocessing is fit_data.csv, after_effects_data.Rdata, and preprocessing.md
• Subset cycling data for use with an After Effects project using to_after_effects.R

Use make preprocess to only run the preprocessing steps and output to after effects.

Preprocessing data visualization and notes can be seen here

• Use random_tree.R to build random trees and random forests using classification on the fit_data.csv file.