Page 1 of 1

Improvement - Multiple overlaid charts

PostPosted: Wed Feb 16, 2011 2:49 pm
by texmurphy
* I would like to see selection of two chart Data types (e.g. Power and Heart Rate) with a possible correlation column.
* A change in the Interval should be detected, currently I have to press chart "Fit to Window" button to see a change.
* Also the chart should be Labeled based on the selection. See mockup below:

Image

Re: Improvement - Multiple overlaid charts

PostPosted: Wed Feb 16, 2011 3:59 pm
by PissedOffCil
Interesting idea.

The focus loss of the interval text field seems to have some problems. You can also click the table, chart or another text field to trigger it. I didn't really look into it but it annoyed me so I'll get to it.

Re: Improvement - Multiple overlaid charts

PostPosted: Wed Feb 16, 2011 9:06 pm
by clackerz
I think the multi-chart selection is a great idea.
Also selecting a bar(s) on the graph should highlight the table and visa-versa

Re: Improvement - Multiple overlaid charts

PostPosted: Wed Feb 16, 2011 9:31 pm
by PissedOffCil
clackerz wrote:Also selecting a bar(s) on the graph should highlight the table and visa-versa


viewtopic.php?f=104&t=10052

Re: Improvement - Multiple overlaid charts

PostPosted: Wed Feb 16, 2011 10:13 pm
by clackerz
DOH! :oops:

Re: Improvement - Multiple overlaid charts

PostPosted: Thu Feb 17, 2011 5:36 am
by PissedOffCil
I was thinking this improvement over this morning (can't sleep again dammit :roll: ) and I'm not sure I see the point of overlaying different charts. What would such a feature give you?

Re: Improvement - Multiple overlaid charts

PostPosted: Thu Feb 17, 2011 6:16 pm
by clackerz
Well for me I've found that comparing cadence (zones), with speed and HR there is a cadence range where I'm working less but my avg speed is up - considering the elevation of course.
BTW I found this using your Filtered Stats plugin - very handy indeed!

Re: Improvement - Multiple overlaid charts

PostPosted: Tue Feb 22, 2011 10:11 am
by PissedOffCil
clackerz wrote:Well for me I've found that comparing cadence (zones), with speed and HR there is a cadence range where I'm working less but my avg speed is up - considering the elevation of course.
BTW I found this using your Filtered Stats plugin - very handy indeed!


I see, but having the distribution overlaid doesn't give you any correlation between the data...


P.S. I also find that certain cadences give me better power output for HR. The problem is that my self selected cadence if 10-15 RPM higher than my optimal power generating cadence. I need to work on re-establishing my self selected cadence.

Re: Improvement - Multiple overlaid charts

PostPosted: Wed Sep 18, 2013 2:27 pm
by rogerpgvg
Could I make a slightly different suggestion? I think it would be very useful if a single distribution graph could show the distribution of two activities (or two sets of activities) for one data track (e.g., speed). I mean, it would look something like this:

Image

For example, let's say I have run two marathons, then by comparing the distribution of speed (or heart rate, cadence or power), I could immediately see whether my speed (or heart rate, cadence or power) in one marathon was higher across the whole distribution or whether I had a few bursts of very high speed that made this marathon faster than the other one.

If you could compare two sets of activities, then you could also compare your training before the two races in more detail by checking how the distribution of speed/heart rate/cadence/power in the activities before the races compared.

If you could compare sets of activities, then it would be even nicer if you could see the two distributions not only in percentages but also in absolute numbers of distance or time. For example, if you have the distribution in percentages, then you might see that your speed in both activity sets 1 and 2 was between 14 and 15 km/h for 10%. But in terms of distance, those 10% may be 50 km in activity set 1 and 20 km in activity set 2 (because you covered more distance overall in set 1). Or in terms of time, 10% may be 5 hours in activity set 1 and 2 hours in set 2. So by seeing the absolute numbers in the distribution graph, you'd get very useful additional information.

Does that make sense, and does it sound like something that could be implemented?