Meteotemplate Forum

It seems that the math on the past two years AVG is off. Anybody else having this problem. I have not verified if this is happening multiple instances (different days). Verifying the data using other plugins, the error does not match the plugin results.

Could you specify a bit more what exactly is happeniing?

Under the heading of 13 Sept/AVG (2 years), the Avg value is off for unknown reason. All other values in the Block are aligned with the Plugin values. This was the case yesterday also, so not date specific. The avg cannot be lower than the min..

Nice catch Bert

Fixed - Upcoming 1.1

Thanks Jachym, works great now.

Jachym,
The Avg calculation is fixed but in the process the min/max numbers for two and three years are the same for today and tomorrow. www/s-p.me (bottom of homepage)

Its not doing this on my page and given you only have 2yrs of data it could be that it simply was the same both last and the year before that.

Jachym, Attached are today's info.
today 14 Sept
Tomorrow 15 Sept
Pix are from the Sept 14 homepage and Yearly stats for 14 Sept and 15 Sept.
There is something wrong here. Maybe its the Plugin?

Looking more into this, it seems that this is a labeling perception problem (yearly averages).
The upcoming info shows:
today avg temp, max temp and min temp (at that time of the day)
Last year same day avg, Max and min. (no problem so far)
then today's date, last year(s) average average(s) but max and min of these years(s)(instead of avg max and min)
The same protocol is applied to tomorrow's info.
The question is: is the multi year data supposed to be all averages of the respective info (avg, min, max) or extremes for the max or min. The title AVG is misleading.

The avg should be average of each year´s avg

Jachym,
Thanks for your reply. In the sections labelled AVG, only the avc value is Average, the others (max and min) show the highest or lowest value for any year, not the average max or min for past years. The info is the same as on the plugin Year Comparison (last line). That line only show the highest or lowest year value, not their average per year.

Jachym,
You state the the averages should be the averages of the max(highs) and min (lows) of each year's max and min averages. I agree with that concept, however that is NOT what is happening. I checked your demo site and the Averages Max and Min are the extreme year averages, not the averages of the max or min years.
If you are using the same data that is on the lower line of the plugin Yearly Comparison, those show the extreme Max Year and same for the Min showing the extreme Min year. In the plugin that maybe the intent as there is no heading for that line of info indicating that they should be max or min averages or extremes year.
I hope that this makes my problem a little clearer.

I think I get it, so what do you propose. Changing the overall max from maximum max to average max?

Jachym,

I think that would be more useful then the extreme max or min. Since we are talking about averages, lets stick with that line. It is in the "Upcoming" block, thus has a forecast tone to it. The block " Station Averages" has the right analysis.

Jachym,
Sorry to bring this up again. You released a new version of this block and apparently overlooked the discussed suggestions.
A few comments:
Today Section:
1) The first section (left) shows "This Year" It would be better to label this "Today". The limitation of this info is that it is up to this exact time info. It does not correlate to any of the info to the right of this section which is for the full day.
2) The "Last Year" Section is correct.
3) The rightmost "Avg (# Years of data)" is going to be flawed if it includes partial today's info. Would suggest to keep it at the # of complete years (like forecast section) In addition we are back to the prior discussed problem of it showing the minimum and maximum of all years instead of the average minimum and maximum of all years. (the data should be the same as in the Station Average Block)

Forecast (tomorrow) Section
1) Looking at the rightmost section (Avg # years of data) the same comments apply as to the minimum and maximum values. They both should reflect the Average of the years minimum and maximum values and not the lowest and highest of any year (extremes). The title AVG is misleading.

I know your attention to details and statistical expertise which is why I am a little bit surprised at the way thIS data is handled in this Block, and yes, I can delete this Block from my front page also...
Either way is fine with me.
I have enjoyed your fine work on this project from the beginning and will continue to.

Hi Bert,
sorry, no its not like I ignored the comments, its just that this forum and emails... I get literally dozens of suggestions and requests every day and I forgot about this. I didnt read the beginning of this thread recently and I forgot we discussed this. I will address it in some future update

UpcomingBlock: Because the API of darksky.net has been discontinued, the block no longer works. The reason is that there is a “die” command in the upcomingBlock.php file, if the API is not found. Because I wanted the block, I commented out everything related to the prediction in the file upcomingBlock.php. Now the comparison works (without prediction).