EvapoTranspiration chart

[AndyKF650]

Hi Hans

I have been working on my local CUtils site and trying out additional charts from the Charts def examples. Most seem to work but those with Evapo Transpiration start in late November 2020 when I installed a solar cell and run continuously to 11 October 2021 when the data stops.

I have attached a screen shot of the offending chart together with a solar chart showing current data. You can see my solar cell is currently producing data, when it should. Also the daily EVT chart shows consistent data.

Is my lack of chart data down to my site or is there a bug in the CUtils software?

[HansR]

Hi Andy,

That looks a bit strange yes.
First, I think the Temp/EVT chart is unrelated to the solar chart (or it should be).
Could you post your original CutilsCharts.def (where the problem occurs, don't cut/paste, don't single out that chart) and the corresponding logfile?

(this would be easier online because of the debugging possibilities, it could also be in the datafiles (json) it now loooking at what you have becomes a bit more complex, but let's try )

Tonight will not be possible, but tomorrow after noon I'll look at it.

[AndyKF650]

Hi Hans

I have attached the whole utils folder so that you can see what is going on. The last up date of the .def file was 15:34 and the log file to run the website update was 15:35.

One other thought I had was that there was a CUtils update on 10/11 October, not sure if this helps.

Thanks for your help

[HansR]

Hi Andy,

I ran your CutilsCharts.def on my data (you can see the result today on my site) and as you see the missing data effect on the end of the TempEVT chart does not occur.
My conclusion at the moment is that is not a bug in CumulusUtils, it must be data driven.

Could you send me the CUserdataALL.json and your dayfile?

[AndyKF650]

Hi Hans

As requested dayfile and CUserdata files

[AndyKF650]

Hi Hans

I have not done anything to the system in the last 24 hrs other than the usual thrifty upload at night but the graph is working just fine now!!!

I hope that you have not lost too much time trying to work out the problem which is apparently not a problem as per the attached screen shot.

[HansR]

Hi Andy, I checked the data in both the dayfile and the json as produced by CUtils.
I see no temporal discontinuity or other strange things which might break the chart so it remains inexplicable at this moment.

I do know that sometimes a chart anomaly occurs where seemingly the x-axis extends beyond the last date of the scale at which position it then give an abnormal value (visual in the chart). What then helps is to change the range using the date selector at the top of the chart: set the last date one day back. That resets the chart and when you then in the same way go one day forward to the original en-date, that might correct the chart.

Otherwise, I have no solution to this problem, it is not in the code, it is not in the data (both to the level that I can understand and analyse) so afaiac it is in highcharts

Any good suggestion where to look for a possible cause is welcome.

[HansR]

Hi Hans

I have not done anything to the system in the last 24 hrs other than the usual thrifty upload at night but the graph is working just fine now!!!

I hope that you have not lost too much time trying to work out the problem which is apparently not a problem as per the attached screen shot.

Ah...
That is a bit in line with my post I just concocted above.
There are depth in HighCharts which we are only just beginning to fathom...

[AndyKF650]

Well this afternoon I did a reinstall of CMX and CUtils and a upgrade of the RPi. Reloaded the various data, reports, .ini and CUser files to bring my system back up to date cleanly.

After all that the system works just fine with all the charts looking as they should.

Maybe I will not need to change anything much for a week or so until the next upgrade!

Thanks for your help