Polar with R

The idea behind

1. Manually download .tcx and .csv file from the Polar Flow website (preferably to a Dropbox folder)
2. run the Polar.R script (it contains a PolarRead function, or more likely the PolarReadAll function that loops through unprocessed files and PolarReads them.

The output is two .csv files (into a Dropbox subfolder)

• xxx_data.csv that contains all the available info in a nice csv file
• xxx_info.csv that has only one line of important values (max speed, average HR, dist, duration, ...)

The goal

• Get familiar with R working on a topic that I like.
• learn ggplot and plot maps
• If things work, I would like to be able to recreate Endomondo/Polar functionality without the need of internet - and also doing something with the back-up exports I always do.
• learn a bit of Shiny

Done so far

• organizing the files so there is some system (and they are properly = nicely named)

• using Git

• script that reads the files, cleanes them and outputs the things I wanted

• added text to inform what files will be read and what is currently happening

• dont drop cadence, power and stride-length - I now (Feb 2016) have wrist-based cadence for running and walking (shamefully, hiking is now excluded in that Polar update).

• add heart rate zone limits 0. do 93:white Rest

  1. 94-112:gray Warm Up
  2. 113-131:blue Fat Burning
  3. 132-149:green Aerobic
  4. 150-168:yellow Anaerobic
  5. 169-188:red Maximum

  aim: calculate time spent in each zone add colour to the map

xxx_data.csv file

done some plotting (w/google maps)

xxx_info.csv

• merge of all headers into one file ('infotable.df")
• plotting some overview statistics

ToDo

• Individual Plots

  • Different axes (alt, HR, speed, cadence) - not possible with ggplot2?!?

• can I speed up the parsing ?!?

• output to public Dropbox (the infotable copy + some charts?)

• what to do when new sport is added? (Feb2016 - first cross country skiing) - set default color for each sport

• option2 to access dropbox files on the internet, not only those stored on C:\

• individual files

  • smooting the curves (make them look like in Polar Flow) - or at least dont plot zero values for HR (that is nonsense)
  • handling situations when I pause the workout and then remember to resume after 5 minutes into the resumed workout
  • pausing vs time stamp
  • during winter time, 6:30 workout starts at 5:30, in summer 10 am workout starts at 8am when reading from tcx
  • run statistics and charts on _merge.csv files
  • function to plot files (if SWIMMING, only show small chart, if HIKING, show elevation, map, ...) but this information is stored elsewhere (info file) - not anymore :) Feb2016
  • Shiny (chose file and analysis + charts come up) 2nd tab

• multiple files

  • some statistics from that dataframe (personal bests, max values)
  • align sport colors with Endomondo colors (swim=blue, bike=yellow, ...)
  • triathlon related trainings analysis
  • running index trend
  • plots = charts
  • Shiny (total statistics) - 1st tab
  • punch card (when I tend to do sports) - first draft on 2015_11_21, but I only have November data
  • last 90 days (yearly statistics already available, next work is expected in January next year)

• nice to have

  • analyze swim data (scrape from Flow?)
  • SWOLF trend
  • analyze activity data (scrape from Flow?)
  • align with google docs file with weights
  • trainload analysis
  • calories of fat burned
  • heatmap on map (where the workouts started?)
  • strava-like "where do people run the most"
  • how are laps handled? (and when do I actually use them?)