I posted recently about how I got started with weight training in middle school. Although I’ve lost the logs I kept so long ago, I do have notebooks dating back to 1998, mostly from various efforts to lose weight. You can read more about the ups and downs in my weight history here on the JMP blog. I’ve since acquired more data from my baby book and various medical records I’ve requested and have a lifetime weight graph (that process and graph deserves its own post sometime).
As time permits, I am typing up my written training logs so that I can visualize and learn from my workout history. I wish getting all my data back to 1998 out of my notebooks and into a spreadsheet was as simple as OCR’ing the pages in my notebooks, but unfortunately it’s not going to be that easy. I have tried several OCR programs on my handwritten data and had no success getting anything useful out of them.
So instead, I have spent time here and there entering data from my notebooks into Excel and importing the data into JMP. I now have data for 13 of the last 17 Januaries, all of my 2014 workouts and all of my workouts and my 2015 workouts through April. I’ll get through all of the data eventually!
In my first blog, I mentioned a few questions that my data is helping me answer about my workout history. I thought would be interesting to explore those questions with the data I have in JMP. My group at SAS develops JMP and I spend time each day exploring internal data sources with our software. But it is also my tool of choice for looking at my activity and exercise data because it offers such a large range of options for filtering and graphing all of my various data types. I haven’t found any other workout software out there that can compare to what I can do in JMP with Graph Builder, Summary and Tabulate. In fact, I’ll be giving a talk about my workout data analysis project at JMP Discovery 2015 in San Diego this fall. My Fitness and Food blog series on jmp.com focuses on using JMP’s data cleanup and visualization tools on my own data.
How often have I worked out over the years?
With this question, I am asking “how many weight workouts did I do.” I haven’t been very faithful about monitoring my cardio workouts, and I’ve done less and less cardio the past few years. To calculate number of workouts, I need a unique workout name, but I couldn’t use my workout title variable for that, since I often repeated workouts. Instead, I used a formula column in JMP to concatenate the workout title and the date on which I completed it to get a unique identifier.
I used JMP’s Summary platform to count these unique workout names by year. Here is the data for 2014, the only full year of data that I have entered so far. I completed a total of 214 strength training sessions. I used JMP’s Recode platform to group unique workout names by the type of program.
As you can see, I mostly do programs designed by John Barban, including Venus Factor and Adonis Factor workouts. I’ve been doing these workouts since early 2010 and have learned so much from following the teachings of John and Brad Pilon. I highly recommend both their blogs (John’s is here and Brad’s is here) and the Venus blog, which includes regular posts I know from women I have met online through their workout community. Although I initially started out doing Venus workouts, I have since transitioned to mostly using the Adonis workouts because my legs don’t need much work after so many years of heavy training. I’ve been very happy with the results of these programs and can’t recommend them enough. I look at it this way-I don’t fix my own car. Why not leave workout program design to the experts?
Last year I did the most workouts (115) from two different older Adonis programs, Muscle Building Foundations (MBF) and Adonis (ATS). Next, I focused on workouts from the Venus Advanced Strength Series (60 sessions), but also did about 20 sessions drawn from various other Venus workouts like the Venus Circuits, Final Phase (aka phase 2) and phase 3 workouts. I classified other workouts I did, like the occasional body part split, randomly designed circuit, and a density training program I found online into the “Other” category (20 workouts).
It’s interesting to look at how these workouts spread out along the year. The following graph shows the number of sets I completed and the total weight I lifted weekly throughout 2014. I assigned custom markers for rows by workout type. Weeks where I only did one type of workout show a custom marker-female symbol for a Venus workout, male symbol for an Adonis workout, D for a density training program. If I did more than one workout type within the same week, a regular dot marker is shown. As you can see, I had a stretch of time in the second half of 2014 where I was alternating between several workout programs for variety.
You can also see where I had an injury in early July that impacted my training, as reflected in a drop in total weight lifted. I hurt my lower back and this impacted chest training too. I reduced my training for a week and when I returned to the gym, I was only doing bodyweight bench pushups for a few weeks while I got over the injury.
Have I gotten stronger at specific lifts over time, either within the context of a specific workout program or between programs?
I could address in various ways-looking at the mean weight used or max weight used for common exercises by year, for example. I’m going to include a scatterplot graph like this in one of my upcoming JMP blogs. It shows the maximum weight I used for selected exercises I performed in 1999-2000 and 2013-2014. In 1999-2000, I wasn’t training in low rep ranges, so it’s not too surprising to see that my max weight lifted for all exercises was lower than now. There are two exceptions, though-upright rows and lateral raises. I’m using less weight on those exercises than I did in the past because I am trying to focus on working my shoulders without letting other muscles like traps assist, and this means using less weight.
I created the following chart to summarize data for my two most commonly performed dumbbell chest exercises. I used a bar to represent my max weight used for the exercise in a month, and a red line to represent the mean weight used across all sets of that exercise that month. I like this view for seeing at a glance whether I was doing mostly heavy training (max and mean are close) or mostly light training (max and mean are further away from one another).
I bet once I get the rest of my data in, I’ll be able to pinpoint exactly when it was that we purchased an incline bench for our workout room at home! I mainly trained at home for several years so I didn’t have access to an incline bench till we bought one off Craigslist.
See the dip in weight used for flat chest presses in July of 2014? Again, the impact of that lower back injury that affected my ability to do DB chest presses while I recovered from it. Understandably, I’ve been hesitant to lift as heavy as I did in the months leading up to this injury.
What exercises do I do the most often?
To answer this question, I summarized the number of times different exercises appeared in unique workout sessions. I also added a primary body part category to the summarized data. This graph points out some of the most common exercises I did for various body parts, like chest presses, a few types of rows, lateral raises and overhead tricep extensions. I also do a lot of shoulder presses and hammer curls.
Which exercises have appeared in more than one program?
This turns out to give very similar answers as the previous question-most of the same common exercises show up. However, instead of looking at how often I did an exercise, this one simply looks at whether the exercise appeared in a program one or more times. A heatmap turned out to be a useful view, I thought. In the one shown below, I filtered the view to show only back exercises using JMP’s local data filter. I have done hundreds of exercises over the years so trying to show them all at once would be overwhelming! Admittedly, I do tweak programs and exercise choice from time to time, so this graph represents the exercises I actually did, rather than what was specifically written into the program. I recently discovered DB 3 point rows as a back exercise and I really like them, so I’ve been using them in place of other rows (e.g., 1 arm rows with a knee on a bench) in various programs. I am sure they don’t appear in any programs I did prior to 2013, however.
I find this and similar graphs to be useful for thinking about exercise variety. I work out at home, in my office and at the gym, depending on day and schedule. What equipment is available to me on any given day dictates my exercises choices. However, even when I have a variety of choices, I sometimes just stick to my favorite and familiar exercises. When working out at my work gym, I have additional equipment choices-e.g., plate-loaded Hammersmith machines-that I could take advantage of for greater variety. I’m trying to do more of that lately, so perhaps I can increase the diversity of exercises in my workout log over time!
This one is getting kind of long, so I’ll answer my other questions in a part 2 post!