I finally got around to making a Heatmap out of some of my downloaded Veloviewer-strava data. Basically, I wanted to see, at a glance, how my rides looked on a diurnal and annual basis – what time of day and through the year.
I’ve only been using Strava since July 24, 2012 so I have a limited data set to work with, but now that I have the structure in place, I can add to it periodically. I was also home with the little one 1 1/2 – 2 days each week through December so my weekday commuting numbers are pretty inconsistent.
So the image shows you all the 15-minute increments when I was riding, each day, from 7/24/2012 through 1/7/2013. So each row is one day, and each column is a 15 minute increment, starting at midnight on the left and going through the day to 11:45PM at the far right.
Some of you may have seen a similar one I did for CaBi ridership a year or so ago. (That one’s a lot more interesting, IMO)
No surprises for my heatmap. As I go through 2013 I would like to push some rides out towards the boundaries (ie. Really early/really late rides) and fill in the middle (lunchtime Hains Point laps!) Unfortunately, I don’t have a way to breakdown the distance traveled per 15-minute increment other than based on the average speed/distance for that ride…if I figure out some way to do it, I could color code the cells to show when I rode fast vs. slow.
Getting the data from the Strava CSV file into this format was a pain in the ass. I work with electric interval data quite frequently and produce similar graphics for electricity consumption (like this one), but the Strava data took a lot of work to get into that layout. Also, for some reason, for the first 4 months, Strava told me that all my rides started 4 hours later than they actually did! Finally, moving from Excel, to PPT to PDF to Google Docs means the images kept losing resolution, so it doesn’t look great – I’m not getting a gold star for pretty, that’s for sure.
.
