Old 01-25-20, 11:37 AM
  #4385  
mattm
**** that
 
mattm's Avatar
 
Join Date: Dec 2006
Location: CALI
Posts: 15,402
Mentioned: 151 Post(s)
Tagged: 0 Thread(s)
Quoted: 1099 Post(s)
Liked 104 Times in 30 Posts
Originally Posted by hubcyclist
I've been geeking out with trying some programming. I took a Java semester course as an undergrad in 1999, and while I enjoyed it, I wasn't convinced I had could cut it as a computer sci major (I spent so much time trying to fix coding errors). I've done HTML and by extension some Javascript and PHP that is related to it, but only modifying existing code, never creating my own.

Based on some popular opinion, I started with Python recently, and by start I mean getting in over my head with a project as opposed to going step by step through a course lol Over the weekend I've been messing around with Strava's API and basically figuring out how to bulk extract my own data to individual files through code and then graph it, which is a little pointless because there are already a million ways to get and analyze data! But it's been fun, I kind of want to build a whole PMC site for myself, which again is super impractical considering the number of free/paid options which exist. Python is supposedly all the rage for machine learning, though I don't see myself breaking any ground in that area

If anyone has any neat ideas of what to try and do as far as building analysis or stuff to do with Strava data I can try and see if I can figure it out, it'll just take a really long time!
Awesome! Python is a good learning language, although I have to deal with a lot of the not-fun parts of it (unicode, 2/3 interop) at work. TypeScript is popular these days too.

A while ago I had a Raspberry Pi that would pull from my Strava feed and then post the ride (title, map, photos) to a Tumblr page. Which basically made a Tumblr site that was a mirror of my Strava feed.

Another idea I had was similar to yours - to calculate CTL/etc and show it on a "dashboard" of sorts, like some kind of small display hooked up to a Raspberry Pi. But the hard part is TrainingPeaks's API isn't available for "personal use" (I wanted to get the actual TSS values, not guess/calculate them), which I can only assume that means it costs money. But I wanted to show current CTL/trends, temperature, etc.

The Pi opens up some possibilities because you can have a "computer" running your code at all times, and talking to all kinds of hardware, etc.
__________________
cat 1.

my race videos
mattm is offline