in reply to Re: Data averages by time of day
in thread Data averages by time of day

Good thought, but not really what I'm after. I plan on this but just a subset of a bunch of reports that I going to generate to give to a doctor.

UPDATE:

The other half of the problem that is out of scope of this thread is the fact that the Perl script I am writing also extracts that data from the Palm Pilot I use to record my readings

Peter L. Berghold -- Unix Professional
Peter at Berghold dot Net
   Dog trainer, dog agility exhibitor, brewer of fine Belgian style ales. Happiness is a warm, tired, contented dog curled up at your side and a good Belgian ale in your chalice.

Replies are listed 'Best First'.
Re: Re: Re: Data averages by time of day
by mutated (Monk) on Feb 26, 2004 at 15:38 UTC
    So are you looking to group the day into "day parts" of the same size each day, so say call them morning lunch afternoon and evening, and then grouping tests by what group they fall in? or more of the given 3-4 tests as points throughout the day put curved lines between the points so you can "guess" what the glucose reading was at some point between your tests? I would go with the daypart methode, only because whatever you do to try and guess what the reading would have been at a given point I would suspect it will be wrong, because of factors like when/what you ate.
    daN.
[reply]

          So are you looking to group the day into "day parts" of the same size each day,

      I'm actually trying graph the averages in a chart when I'm done. (wish I could put an image in this post!)

      Peter L. Berghold -- Unix Professional
      Peter at Berghold dot Net
         Dog trainer, dog agility exhibitor, brewer of fine Belgian style ales. Happiness is a warm, tired, contented dog curled up at your side and a good Belgian ale in your chalice.
[reply]
        I encountered a related challenge recently, charting blood pressure numbers over the course of several months.

        My conclusion: while averages are nice, full data spreads are even better.

        In addition to your averaging, try just doing a traditional scatter plot: time of day along the x axis, result values along the y axis, and a dot on the chart for every data point.

        Once you've done that, things like the average for some time period or the outlines of the daily curve tend to become visually apparent, even to someone who isn't fluent in statistics.

[reply]

In Section Seekers of Perl Wisdom