A bit of background. WeeWX manages two different sets of databases.

First is the main archive table. It is updated with every archive record.

Then there is a set of "daily summaries." These contain the minimum,
maximum, time of minimum and maximum, and some other details for each
observation type for each day. It is updated with every LOOP packet, so
despite having only one record per day, it can be quite precise about when
minimums and maximums happen. Lookups using daily summaries can also be
much faster, but they require that the aggregation period run precisely
midnight-to-midnight.

On to the problem. When WeeWX evaluates a tag such as $day.outTemp.min, you
are asking for the minimum since midnight, so the daily summaries are used.
This gives an answer down to the resolution of the LOOP packets.

But, for a tag like $span($hour_delta=24), the aggregation period does not
run midnight-to-midnight, so the archive table must be used. So, the
results have a resolution no better than the archive period.

With your dataset, the minimum of 984.9 happened at 10:38:36 and the daily
summaries captured it. But if you were to look in your archive table, it
would show 985.0 at 10:20:00, because that's the average over the 5 minute
archive period.

If you hate this result, you can tell WeeWX not to use the LOOP packets to
update the daily highs and lows by setting option loop_hilo under
[StdArchive] to False. This will cause the daily summaries to be updated
with the min and max archive records, rather than LOOP packets.

Hope this helps.

-tk

On Wed, Oct 21, 2020 at 3:37 AM Mike Thompson <
[email protected]> wrote:

> Hi All,
>
> I'd like to have information over the last 24 hours and I thought I could
> easily achieve this by defining a  $span tag but it's not worked out as I
> thought.
>
> I've defined the tag as $span($hour_delta=24) and as $span($day_delta=1)
> both give the same results.
>
> To test I changed the Seasons statstics.inc to include my 24hrs as follows;
>
> #set $archive_data = [$day,$yesterday,$span($hour_delta=24),
> $week,$seven_day, $month, $year, $alltime]
>
> This seems to work ok as shown below
>
>
>
>
> But there is the occasional anomaly for example looking at the min
> Barometer it shows;
> 984.9 for today @ 10:38:36
> 988.1 for yesterday  @ 09:54.36 and
> 985.0 for last 24 hours @ 10:20
>
> I was expecting the min for last 24 hrs to be equal to or less than
> today's value (984.9 is less than 985) but it's not, why is that?
>
> It doesn't just happen with min barometer it is the same with max temp
> which was yesterday's example.
>
> Appreciate your thoughts
> Thanks in Advance
> Mike
>
> extract from Seasons statistics.inc showing min Barometer
> <tr>
>           <td class="label">$obs.label.barometer</td>
>           <td class="units">$unit.label.barometer</td>
>           #for $archive in $archive_data
>           <td class="data
> new_row">$archive.barometer.max.format(add_label=False)<br/>
>             <span class="timestamp">$archive.barometer.maxtime</span><br/>
>             $archive.barometer.min.format(add_label=False)<br/>
>             <span class="timestamp">$archive.barometer.mintime</span>
>           </td>
>           #end for
>
> --
> 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/69f8d05a-f536-4be2-9b89-468684983cf5o%40googlegroups.com
> <https://groups.google.com/d/msgid/weewx-user/69f8d05a-f536-4be2-9b89-468684983cf5o%40googlegroups.com?utm_medium=email&utm_source=footer>
> .
>

-- 
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/CAPq0zECNBwnrsjVS%3D-L5cWMfgBtMheRU1LrrAou5csOvf%3Dc%2B_g%40mail.gmail.com.

Reply via email to