My training tree

I made a little "toy" of my training log in 2012.  The idea was to take my entire log and use the information to be able to predict how I trained given two bits of information - the day of the week and the activity.  Here's how it works:

1. The top of the tree begins with the day of the week.  Today is Tuesday, so the top of the tree means that I branch off to the left (if today was Saturday or Sunday, I'd branch off to the right).
2. The top left branch gives activities.  I rode my bike, so that's "cycling" which means I branch off to the left.  
3. The left branch gives a number - 46.97.  That's the estimate of the minutes that I trained.  

And that is, in fact, almost exactly right.  My training for today was 45 minutes of biking on my way to and from work.

Here's another example:

1. Let's say I want to predict my training for Saturday.  From the top of the tree, I go to the right branch.
2. I'm going to go orienteering.  So the activity is orienteering and I go to the left branch.
3. The left branch gives a prediction that I will orienteer for 54.75 minutes.

One last example:

1. Let's say that in addition to my bike riding to work, I am going to go for a run today.  Since it is Tuesday, I begin at the top of the tree and take the left branch.
2. The left branch is activities and my activity is running.  That means I move down to the right branch.
3. That branch asks me for the day of the week.  Since it is Tuesday, I move down and to the left.
4. The left branch gives a prediction that I will run for 27.39 minutes.

I think it is an interesting way of looking back at a year of training.

I should play around with the data a bit more.  I think it might be good to look at the season (probably the month of the year) as well as the activity and day of the week.

Back to okansas.blogspot.com.