Chart Problems when Changing Axes (Orientation, Direction, Horizontal/Values)

Suggestion Box Bugs
1

I seem to have a problem with my charts.

When I select a certain horizontal axis and vertical axis it works fine:

image
image1468×812 35.3 KB

But when I swap them it doesn’t display anything.

image
image1474×804 5.03 KB

The settings that trigger the problem:
image
image676×652 12.4 KB

What’s really weird is that deleting one row (which brings the table size down from 1001 to 1000 rows) enables the chart again. It doesn’t seem to matter which row I delete.

image
image1444×800 11.7 KB

I’ve tried many different duplicate handling functions it doesn’t seem to matter which one I use (sum, median, mean, mode).

Is there a chart size limit? (if so could you please add a warning so that I know what’s going on?)

Why does this error only apply to one axis orientation and not both?

Why does my chart look so different when I orient it this way? Is there no way to do index reduction for this axis orientation?

If not, then how do I plot multiple Percent of Time lines on the same chart?

2

Hi @cnr,

I did some testing, and yes, it does appear that we have a hard limit of 1,000 rows for a line chart. Honestly, I don’t think we have a warning because as far as I know, this is the first time someone has hit that limit.

For the step down line, can you check in the table chart settings and under “Layout”, there is a “Smooth Line” option you can turn on that might help.

Our charts are due for some added functionality. There are some current limitations that we are logging and hope to remedy when they pop up to the top of the list.

As for why you’re not seeing anything when you flip the axis, we might need to see the doc to test that out. My data set isn’t showing the same issue. If you want to copy the columns to a new doc so it’s generic data at that point, then you can share it here and we can all dig deeper.

3

Thank you so much for the response!

Changing to smooth lines doesn’t help because of the structure of the data

Here is a link

4

I know it’s not a general solution, but it looks like you can reduce the number of data points easily, since Percent of time is the same for a range of Occupancy. To keep the steps, you only need to draw a data point before the step down and after the step down for each value change.

E.g. here I reduced the number of rows from 1001 to 62 while still keeping the same stepped line:

To remove the steps and draw a smooth line, just remove the OR Is step before… from the filter on the chart. And to restore to the buggy behavior with 1001 rows, remove the filter altogether.

4 Likes
5

Cool! Thank you for the help Paul, that’s a good trick. Since I’m trying to do all this with multiple different tables, it would be nice to eventually have this built into the product (just as there are averaging tools built directly into the charting for the other direction). For now, I’ll probably just do this data cleanup in the script that outputs the data, that way I don’t have to do it in Coda for every table.

Thank you!

6

Actually, another question: would this work if there were multiple Percent of Time columns? For some reason I feel like it would not work, though I have not yet played with it

7

Why wouldn’t it? :slight_smile: and why won’t you try?

If you need another Percent of Time column for the same monotonous 1001 samples of Occupancy, then all you need is update the Is new step calculation to detect new steps from another column to plot more data points, I think.

(besides, realized I could make it much more optimized, the calculations)

8

I didn’t have the time at that moment to play with it, but yes! it does work! I’ve also decided to reduce the number of rows (so I have less samples) which should improve the performance.