It would be extremely nice to be able to configure the y-scale on charts for any given workout. I have an activity where I gained ~100ft elevation and lost about the same, yet it looks like a straight line because it is scaled 0-1200ft.
You can take your mouse pointer and drag out a selection window on the graph. Frequently I can get the scaling I need by doing this. The issue I think for GC is how do you adequately scale something that usually over 35 miles (184,800 feet) long by only a few hundred feet high and show it in a small rectangle on the page.
The programming to auto-scale ordinates between maximum and minimum (using recorded data contained in file) is not that difficult. I am not a programmer; I didn't play a programmer on TV; and I didn't stay in a Holiday Inn; but, I've done enough programming--including scaling data for graphing--to be certain that it's not that difficult.
I agree that it's not too hard. One of my first assignments out of school 25 years ago was to write software to plot test data. I wrote a little routine to automatically scale the axes of the plots. I knew just enough programming to be an engineer, so if I can do it, it's not too hard. And in some ways, scaling the graphs in GC would be easier, because some assumptions can be made regarding things like elevation and pace. The primary problem I see is that the scale can be overly influenced by just a few bad data, which often happens at the beginning of an activity (e.g., pace hasn't stabilized yet, or the initial pace is infinite because you haven't started moving yet). Instead of scaling the graph to the min and max of the activity, these spurious points could be filtered out, which would be pretty simple to do.
but on a graph that trying to represent 35 miles (184800 feet) or more across the x-axis, and only a deviation of 300 feet across the y axis for most of us, it's going to look pretty spikey or exaggerated when put in that window on GC that's only 6 inches by 2 inches on my display. Dragging out a selection box still works well for me.
but on a graph that trying to represent 35 miles (184800 feet) or more across the x-axis, and only a deviation of 300 feet across the y axis for most of us, it's going to look pretty spikey or exaggerated when put in that window on GC that's only 6 inches by 2 inches on my display. Dragging out a selection box still works well for me.
This should not be an issue. The scale for the y and x axes are independent. In your case, the y-axis should go from 0 to 300 feet, and the x-axis should go from 0 to 35 miles. Or maybe I'm missing your point?
My issue is that during a run, the pace scale will sometimes be from something like 60:00 minutes per mile to 0:00 minutes per mile, even though my pace while running was between 8:30 and 9:30 per mile. Look at the attached image for an example of this. That makes the graph completely unusable. Same thing for the elevation. Look at the attached image for a run I did last week. According to GC, my elevation ranged from 78 to 127 feet above sea level, yet the scale on the graph ranges from -200 to +200 feet.
Well I do have to agree with you on the scaling for pace. But on elevations I still like the 'softened' picture I get for the overall view as it's done now. It seems like GC or some other site I used before used to plot the y axis scale to the min and max elevation in the activity and then it looked so spikey that it didn't give a good overall picture of the ride profile to me.
Well I do have to agree with you on the scaling for pace. But on elevations I still like the 'softened' picture I get for the overall view as it's done now. It seems like GC or some other site I used before used to plot the y axis scale to the min and max elevation in the activity and then it looked so spikey that it didn't give a good overall picture of the ride profile to me.
You make a good point about the softening of the elevation curve. I think the solution would be that the scale should pretty much follow the ride/run max/mins but there should/could be a 'smoothing' checkbox that tidied things up a bit.
You make a good point about the softening of the elevation curve. I think the solution would be that the scale should pretty much follow the ride/run max/mins but there should/could be a 'smoothing' checkbox that tidied things up a bit.
The only way to soften it is to adjust the scale of the y-axis which whether intentional or not is what GC does right now. If you adjust the scale of the y-axis to only the range between the min and max elevation it's going to be like looking at a seismometer of a magnitude 8 or 9 earthquake.
I'm only saying this for the elevation graph because by leaving the scale of the x and y similar, you get a reasonable profile of your ride. If you go through the math to plot these points, you'd quickly see what I'm talking about. Remember you are trying to show this data in roughly a 2 inch x 6 inch box, at least that's about the size on my screen.
I'm only saying this for the elevation graph because by leaving the scale of the x and y similar, you get a reasonable profile of your ride. If you go through the math to plot these points, you'd quickly see what I'm talking about. Remember you are trying to show this data in roughly a 2 inch x 6 inch box, at least that's about the size on my screen.
Fair enough.
I'm happy that the profile looks a bit exaggerated to be able to pick up the finer detail. Scaling to have x and y similar, results in an almost totally flat profile. e.g Here's my (short) ride home yesterday.
the ratio of scale is approx 25:1 (So 25 times exaggerated)
And here's how it looks if I drag a box over the profile
the ratio of scale is approx 50:1 (So 50 times exaggerated but more detail)
I'm going to have a play with a longer ride to see how that works out.
