Extending Charts in CUtils

[AndyKF650]

Not doing very well this afternoon!!

Try again Andrew!

[HansR]

In CutilsCharts.def change the first two occurrences of SolarRadiation to MomentarilySolarRad, that should do it.

And in future that will be CurrentSolarRad but not for the executable you have.

[AndyKF650]

Hi Hans

Well that is starting to make sense, the Overview and recent solar charts look just fine now.

The long term ALL Solar is starting to work but the final entry looks very odd as it contains a sum of all values when I looked at it after a CompileOnlyy . Then I ran the full Website update and the chart looks just great as per the attached screen shot.

The current log file and charts .def files are attached.

I suppose the next movement is for the ALL Temp chart which still just shows loading

[HansR]

OK, working on it besides other things ...

[HansR]

OK. Something totally different: https://assets.highcharts.com/errors/18/
Tomorrow is another day

[AndyKF650]

Hi Hans

I have been doing some testing this morning on various keywords and eval statements. The first test was just to have ALL Temp which worked just fine, one complete record graphed.

Then I tried various iterations of ALL Temp and Average Dewpoint just changing the spelling of Dewpoint in the expression EVAL [ (MaxDewPoint + MinDewPoint / 2) ] I was unable to get any combination to work. I also did a similar exercise on FeelsLike to no success.

Any other thoughts from the master

[HansR]

Hi Andy,
You're becoming a real tester putting the compiler (me?) on the rack
OK.... but you really need to enclose or embed the corresponding CutilsCharts.def (or multiple if it is a combination issue) otherwise I can't make much of it.
Regards,

[AndyKF650]

Hi Hans

Following on from previous testing I have attached the current .def file and the last 2 log files. You will see that I have put anEVAL expression for the AverageFeelsLike at the top of the .def file. I tried to see if doing the calculation and then putting the definition without the EVAL expression on the same line would work. It did not!

Looking at the ...929 log file it seems that certain keywords are not being found for the failing charts, have these changed against those I use here.

As an extra I noticed today that the ALL pwsFWI chart has stopped behaving normally.

[HansR]

Hi Andy,

OK. Thanks.
The most important is the charts code, more than the logs as I use that for reproducing. So if you say a chart works on its own but in combination with others it does not, then just post the code of the combination. I run it myself and see what comes out.
If only two charts show the issue, then leave the rest out and post the minimal charts code. Reduction is the key to success here.

Oh, and you write you try a lot of spelling exercises: that is not really necessary because CDL is case independent. I mostly use (Pascal) camel case for variables and capital for keywords because I like that style and makes it easier to read but it does not matter: the parser compares case independent.

Regards,

[HansR]

Hi Andy,

Patch 2 is online in the distribution post.

The issue was that if you did not specify an axis for an equation it did not get one because you do not know what dimension the outcome is (so, an axis was actually implicitly obligatory which is kind of a specification error). Now, if you do not specify an Axis for an equation, the FREE axis is assigned and a warning is written to the log. I think all other averages are solved with this one.

Do you have any others

Regards,

[AndyKF650]

Hi Hans

Well I downloaded the patch 2 script and loaded, and ran it through the usual update process. No problems at the initial run but the Solar Charts did not work, this was down to the MomentarilySolarRadiation being used rather than CurrentSolarRadiation. I changed the definition file and that worked fine

Next moving onto the pwsFireindex chart no issues there now.

Moving onto the ALLTemp chart this did not work. I tried various fixes to get it to compile and did eventually succeed, as per the attached .def file however running the website update produced a Loading.... notification rather than a real chart

It is strange that the log files on compileonly and website seem to indicate that it is all good but in reality it does not work.

[HansR]

Moving onto the ALLTemp chart this did not work. I tried various fixes to get it to compile and did eventually succeed, as per the attached .def file however running the website update produced a Loading.... notification rather than a real chart

It is strange that the log files on compileonly and website seem to indicate that it is all good but in reality it does not work.

That is a nice find thank you

I had just stated that CDL is case independent so you write:

Code: Select all

( MaxDewpoint + MinDewPoint ) / 2

The subtle difference between the two
Indeed I had two calls where there was no case independ comparison. I will not bore you with the details but you can find issue and solution here. A bit technical and theoretical.

Anyway, I have to follow the flow to check a bit more, then patch 3 will follow.
Great testing !

[AndyKF650]

Ah Ha the mist lifts!!!

The All Temp chart is now starting to work with Average temperature and average feelslike showing as lines in a chart

The odd thing is that average dew point still does not show despite the correction in my spelling

The current .def file and logfiles are attached

[AndyKF650]

If you define the Axis as Temp the chart looks even better but still no dewpoint

[HansR]

My indication of the subtlety in

Code: Select all

( MaxDewpoint + MinDewPoint ) / 2

made you change it to

Code: Select all

( MaxDewPoint + MinDewPoint ) / 2

but it must be

Code: Select all

( MaxDewpoint + MinDewpoint ) / 2

With issues like this you really need to read the spec. of the ALL/DAILY Plotvariables.
But anyway it is for the 'emperors beard' because with patch 3 all variations, even all lower case or upper case should work.