Thanks Gary I will try your extension.
The instructions say to do this:
wget -P /var/tmp 
https://github.com/gjr80/weewx-gw1000/releases/download/v0.1.0/xcum-0.1.0.tar.gz
The above is not found .
Should be below statement:?
wget -P /var/tmp 
https://github.com/gjr80/weewx-xcumulative/releases/download/v0.1.0/xcum-0.1.0.tar.gz

Question: If I install it do I need to do anything else to produce the 
cumulative graph?
Do I need to remove anything that is already there in the config file?

I don't want to break my station :)


On Sunday, 25 June 2023 at 05:35:35 UTC+10 gjr80 wrote:

> The current WeeWX cumulative series xtype is rather primitive and only 
> produces a series where the first data point is zero and subsequent data 
> points are cumulated by the aggregate interval over the plot time span. The 
> default day, week, month, year plots produced by the image generator cover 
> a (roughly) day, week, month and year from the current time. For example, 
> the day plot starts on a three hour boundary 27 hours before the next three 
> hour boundary after the current time (eg it is 7:30pm as I write this so a 
> default day plot produced now will cover the period from 6:00pm yesterday 
> until 9:00pm tonight). So what? Well if the plot was cumulative the first 
> data point (6:00pm yesterday) would be zero and the plot would cumulate 
> from there. No reset at midnight (or any other time for that matter). 
> Similar effect for week, month and year plots. So a year plot produced 
> today 24 June 2023 would have its first data point on 1 June 2022 (year 
> plots start on the 1st of a month) and it would be zero, the plot then 
> cumulates from that point on until today, 24 June 2023. No reset on 1 
> January 2023. In other words the value of the last point of the plot will 
> include data from part of last year as well as all of this year.
>
> So the effect your are seeing is expected. Looking at your site the day, 
> week and month plots indeed look correct, not because of the cumulative 
> aggregate behaviour but rather because there was no rain in the plot period 
> before midnight, Sunday and 1 June on the day, week and month plots 
> respectively. If you wait until July you will note you month cumulative 
> plot will be out because it will include rain from June. In time your 
> cumulative day and week plots will show similar 'wrong' data.
>
> What to do about it? You either ensure your plots start on the time that 
> you want the cumulative value to reset (eg midnight for day plots) or you 
> change the cumulative series xtype behaviour. Neither can easily be done by 
> the user via config changes. The former is just going to produce ugly plots 
> and will be difficult/messy to implement, modifying the cumulative series 
> xtype has the advantage of needing no change to the WeeWX plot engine. Late 
> last year it got to me and I produced a WeeWX service that replaces the 
> cumulative series xtype shipped with WeeWX with a more capable version that 
> allows the user to specify a reset time for the cumulative value (the 
> service works by placing the more capable cumulative series xtype ahead of 
> the shipped cumulative series xtype in the xtype list so the modified 
> version is always used). I've been using the modified xtype with my solar 
> PV system to report cumulative daily energy since late last year and it 
> seems to work acceptably. I haven't gotten around to implementing it for my 
> (weather) station.
>
> If anyone is interested you can find it on GitHub here 
> <https://github.com/gjr80/weewx-xcumulative>. The wiki has some basic 
> details of how to configure the image generator plots to use the reset 
> times as well as some example plots. Note the extension requires WeeWX 
> v4.6.0 or later.
>
> Gary
>

-- 
You received this message because you are subscribed to the Google Groups 
"weewx-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/weewx-user/d23a8e0b-62c2-444f-b061-d316bb5df337n%40googlegroups.com.

Reply via email to