a) you should use an interceptor driver
version which cleanly handles all your sensors - and also those
you don't have
the original version on GitHub was never extended and is a
crippled version from a today's point of view as it covers the
basic sensors only and its exclusion field-map is incomplete -
doesn't match the current customized server protocol sent by a
modern Ecowitt console with more than the basic sensors
connected - extended, complete and working versions of
interceptor.py and extensions.py (for units) exist though
if you use the device_type ecowitt-client as you already found
out yourself, the assignment of the rain data should work
properly
But that's "beauty talk"
b) what hits my eye is your archiving
interval of 15 seconds ...
Nov 1 09:04:25 goatbox weewx[281767] INFO weewx.engine: Using archive interval of 15 seconds (software record generation)
I
have no idea what added value you try to get from such a short
interval and it may overload the system and be - at least
partially - responsible for your issue.
--
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 weewx-user+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/weewx-user/433cd173-d2f6-466e-9f16-983ef4e772e1n%40googlegroups.com.
I think you have to understand better how
weewx works:
even if you set a archive interval of 300 seconds (5 minutes),
each data packet sent by your console every 16 seconds (shorter
doesn't make sense for you anyway as the outdoor sensor array
only transmits every 16 seconds) is processed by weewx and added
to its Loop data according to the accumulator setup. The "loop"
is a table in which all data which occur during an archive
interval are collected [summarized, averaged etc.] and the last
version is then archived at the end of the archiving interval.
So nothing gets lost and you don't need to archive every 16
seconds ...
You can find a proper interceptor.py and
extensions.py at
https://pc-wetterstation.de/weewx/usr/share/weewx/user/interceptor.zip
and
https://pc-wetterstation.de/weewx/usr/share/weewx/user/extensions.zip
To view this discussion on the web visit https://groups.google.com/d/msgid/weewx-user/55b97aba-2533-4be3-8c26-b00593da4155n%40googlegroups.com.
And, to improve your knowledge of your
weather station, I recommend reading the WiKi at
https://www.wetterstationsforum.info/wiki/doku.php?id=wiki:wetterstationen:ecowitt-stationen
if you understand German (the Froggit brand may point to that).
If not, try https://www.wxforum.net/index.php?topic=40730.0 in
English - not as extensive but the best on the web in English
language
by the same author
1. as Gary (@gjr80) already mentioned - send us your import.conf file for the CSV import and maybe 5 lines of your import CSV file including the first line with the descriptors
2. did you run the daily summary update
after the import ?
sudo wee_database ...... from= to= (see wee_database --help)
3. did you compare the values from the
import CSV file with what was written to the database ?
(download/install sqlitebrowser unless already installed and
start it with "sudo sqlitebroswer /var/lib/weewx/weewx.sdb"
(or whatever is the path to your weewx database) and compare the
added records:
convert your date/time stamp into EPOCH to find the
corresponding record in the first database column)
To view this discussion on the web visit https://groups.google.com/d/msgid/weewx-user/d8fc9c17-7347-4e16-8b59-620a9c28c62en%40googlegroups.com.
1. as Gary (@gjr80) already mentioned - send us your import.conf file for the CSV import and maybe 5 lines of your import CSV file including the first line with the descriptors
2. did you run the daily summary update after the import ?
sudo wee_database ...... from= to= (see wee_database --help)
can you remove all lines with
";;;;;;;;;;;;;;" from the import CSV file and start the import
again (I assume you still have an untouched copy of your
database),
re-create the NOAA reports and check again ...
The NOAA reports from your web site is the only place where I
could compare the import file content with your web site
database.
Obviously at certain days no data or not all data of that day
were imported.
Can you also provide the syslog from this import (or from your
last import) - maybe for certain timestamps already a record
existed and wee_import couldn't create this record again.
wee_import only creates new records - doesn't update existing
records !
--
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 weewx-user+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/weewx-user/86c605c3-1b72-4c79-aad4-ff18bb97a7dcn%40googlegroups.com.
I have no idea why it should be 800+ mm
I did the work deleting all 0.0 mm value records and left for
non 0.0 days only the last record (usually either 18:00 or 22:00
whatever was available)
and added them all up - result: 507.2 mm as opposite to the
470.2 mm shown in your graph and NOAA. Not 800+
PRECIPITATION (mm) MAX ---DAYS OF RAIN--- OBS. OVER YR MO TOTAL DAY DATE 0.30 3.00 30.00 ------------------------------------------------ 2023 01 21.6 17.3 05 2 2 0 2023 02 6.3 2.8 20 6 0 0 2023 03 100.8 56.4 05 7 4 2 2023 04 13.0 4.6 22 5 3 0 2023 05 111.2 56.9 31 13 7 1 2023 06 79.4 22.6 15 15 5 0 2023 07 35.9 14.7 09 12 3 0 2023 08 6.6 2.0 23 5 0 0 2023 09 68.3 48.5 25 7 3 1 2023 10 27.1 11.4 21 5 4 0 2023 11 0.0 0.0 01 0 0 0 2023 12 ------------------------------------------------ 470.2 56.9 May 77 31 4 also, your NOAA reports are not consistent - e.g. sum Oct 23 in the yearly view: 27.1, in the Oct 23 monthly view: 23.9
Going through the import field map and the
import CSV file,I cannot see any reason where this (meanwhile
as far as I could tell in a first rough check) only 35 mm
difference comes from.
All these consistency checks you could have done yourself and
come to the same results as I did, couldn't you ?
I attach the Excel file I created from your import.csv - the
other columns are hidden (mark adjacent columns --> Format
--> Hide/Unhide)
No guarantee I missed some lines - but you can repeat the
approach - and make sure all your NOAA reports are consistent
amongst each other in the first place.
As a practical/pragmatic approach I suggest you complete the
Excel for each month - create an additional import file with a
time stamp one minute later than the last one each month (using
the other data from the last record) with the missing amount and
import it
Or use sqlitebrowser as suggested earlier to add the missing
values of the respective days to the day already imported.
When you use sqlitebrowser, you will need to update the daily
summaries (wee_database)
In the end, at least you will have a consistent database.
If Gary or someone else has the time to analyze the data more
deeply to figure out why wee_import does an incomplete job, they
shall feel free to do so
To view this discussion on the web visit https://groups.google.com/d/msgid/weewx-user/06c2d1f3-4972-4263-9084-085af01f890cn%40googlegroups.com.
summary data shown on the ecowitt.net dashboard are the values from the console and can be changed by the user in the console. At ecowitt.net no summarizing is done for the displayed tiles, just display of what the console sends. What is shown there is user responsibility. If the user makes adjustments, they are simply taken over and displayed. => you can only conditionally expect accuracy in the summary counts - it may or may not not match the really stored 5-minute (or longer period summarized data depending on the retention policy) data. (we can see the discrepancies between import summaries, NOAA summaries and dashboard tile summaries)
If there is a discrepancy between this
summary data shown on the dashboard tiles and the summarized
end-of-day data for rain, this is most likely user caused. (800+
vs. 470)
Data from the import was all properly stored - discrepancies are
likely to be user caused and not we_import caused.
--
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 weewx-user+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/weewx-user/dc17cabc-7722-4a27-ac6b-5f96ec306786n%40googlegroups.com.
Hi Marlon
three more remarks:
1. comparing to neighbouring environment may not always be
helpful - as rainfall is not a homogeneous phenomenon and 100 m
away from your sensor there may even be no rainfall at all - or
vice versa. A proper manual gauge close to your sensor makes
more sense.
2. your WH65 arrray with its rain gauge may
not be at the same height as your manual gauge ! To be
considered.
Did you clean your rain gauge (WH65) properly ?
3. if you are looking for more accuracy:
get yourself a WH40 with a rim height extension
(https://www.wxforum.net/index.php?topic=44074.0) and a GW1100
(or GW2000) as your WH2910 console (Froggit WH3000SE) cannot
receive the WH40 - mount it at 1-1.5 m above ground and put your
manual gauge close to it - same height.
Use a calibration factor of 1.02 for the WH40 - due to
imperial/metric conversion.
And then you are much better set for your endeavours.
Your issue(s) is(are) in my opinion not weewx related but
console/sensor/location related.
To view this discussion on the web visit https://groups.google.com/d/msgid/weewx-user/d4599cf2-5d71-4c5d-9c25-3d0b5c7a35e3n%40googlegroups.com.