As promised, version 2 now offers much better grounds for further development so here is another plugin, related directly to your station data again.
- comparison of data using a polar chart
- data break down by years and months
- overall normals
- summary table
- table with deviations from normal
Unzip, paste to plugins folder, add link to menu 🙂 Cannot be simpler 😉
Note: the plugin should still be fully compatible with versions 1.x though I do recommend upgrading to v2.0 Watermelon
10 thoughts on “New Plugin – Annual Normals”
I knew that kind of means you did not like
well it is not that i “dont like them” – the problem is that if I did that the data will be very inaccurate, so you should not do it too.
I do not know if I understood well.
the months of May, June and July have complete records of four years (2012-2015). It is true that in 2012, the frequency of the data was every minute; in the other years is 5 minutes).
Assuming that was the case, I could change the overall average?.
That is, can you do the simple average of the months?
I say this because I want to know the normal values of my station
Yes, but that is exactly why I dont want to change it Javier, because if I did it would be inaccurate.
Think about this:
It is 1st of the month, so there is just one day from that month for that year. But, if I did average of the years, for this year it would take the value from just this one day, so it would be biased.
Imagine for example that it is 1st December and it was very warm. But then, for December 2015 it would have only this one day, so it would show a very high value and if I did what you suggest it would take this one day´s value as whole month´s average and all the numbers would be biased.
I installed the plugin “Year Normals” which I find very useful:
Curiously I have observed that monthly averages are correct, but the overall averages are not all correct.
Admitting that the rounding can give a deviation of 0.1, it has struck me that there are four months with major deviations.
(in parenthesis I write the right averages):
Total Monthly Average temperature:
January: 10.3 ° C (10.3) error = 0
February: 8.5 ° C (8.8) error = -0.3
March: 11.7 ° C (11.8) error = -0.1
April: 13 ° C (13.1) error = -0.1
May: 13.8 ° C (14.9) error = -1.1
June: 17.1 ° C (18.2) error = -1.1
July: 21.3 ° C (20.6) error = 0.7
August: 20.6 ° C (20.8) error = -0.2
September: 19.6 ° C (19.4) error = 0.2
October: 18.3 ° C (18) error = 0.3
November: 14.9 ° C (14.8) error = 0.1
December: 10.4 ° C (10.5) error = -0.1
He commented as a curiosity and not as a criticism
this could be due to the fact that you might have different number of records each year.
The overall average is not calculated as average value of the individual years (which is what you did), but it is calculated by really taking all the values for that month and dividing by their count. Because if for example one year you have more records than other, it would not be accurate if I just took the average from the annual averages. But because probably the difference in the number of records each year is very small, also the differences are very small.
I have already installed the new update of the template.
Many improvements are appreciated.
From San Sebastian, Spain:
Thank Jachym! and congratulations for your great creation
Thanks Jachym I tru to install now. Can I do a future request ? I should like to see the yearly totals too and maybe seasons comparisions (the lates must be hard I think..)
Download not found on this server 😉
Sorry about that, should be ok now.