Studio711

Weight Loss Data

Yesterday I wrote about the process of losing weight, but I only alluded to data sets. Let’s dig in! My data sources were the Fitbit API (for both number of steps and weight) and some weather history which I’ll explain later. This all came into through Power Query.

First of all, here is a chart showing my weight over time and then which days of the week. I usually gained weight on Sundays (our days to get together with family), Tuesdays (taco day in the café at work) and Fridays (the day we sometimes go out to eat.)

Next up are some charts showing the number of steps I took each day. The chart on the left shows how often I took a certain number of steps. The chart on the right shows the total number of steps each month. Note that we’re only halfway through June so that bar is shorter.

I fully expected to see a correlation between the number of steps I took on a given day and the amount of weight that I lost. Nope. Here’s a scatter plot showing no correlation. I think that walking is good for weight loss if you’re very overweight and you don’t move much. But there’s a point where walking is just too efficient to do much additional good.

I then started looking for other possible data correlation. Maybe the number of steps that I took was related to the temperature? Nope.

We’re in a wet part of the country so maybe the amount of rain we get in a day dictates how many steps I take? Not really. My really big days have happened when it doesn’t rain, but just because it’s dry doesn’t mean I’ll walk a lot.

In the end, I took all the various data points and ran them through Excels correlation algorithm. Nothing came out showing any real correlation. The biggest one was one of the charts you see above: the bigger the high temperature, the more steps I take, and even that was only a 0.48 correlation. That’s skewed quite a bit too because I’ve been doing a LOT more yard work lately and it has been warmer.

Even though I didn’t find a scientific way to lose weight, I did learn lots of things that AREN’T related and that’s interesting too!