Though I posted this yesterday but I wasn't able to see it. Anyway cleaned up the records I updated with NUL and put in the value 0.0 for the radiation fields and after a few runs of Belchertown the record high of Solar is now gone. The NUL value was the root of the issue.
On Monday, June 9, 2025 at 10:00:47 AM UTC-5 vince wrote: > I cleaned mine up yesterday as I suggested and Belchertown records caught > up just fine. > > On Monday, June 9, 2025 at 4:00:51 AM UTC-7 RonV42 wrote: > >> Thanks I will dig deeper into the archive as you suggested I have done >> cleanup many times with false rain and wind and use "zeros' for the value. >> I am going to try setting zeros around the time where the station was >> reporting excessive radiation. >> >> On Saturday, June 7, 2025 at 10:29:19 AM UTC-5 vince wrote: >> >>> Quick glance seems to say it is displaying $alltime.radiation.max and >>> $alltime.radiation.maxtime.raw like any other skin would, so it doesn't >>> seem like Belchertown does anything special in how it displays records. >>> >>> Check your db again. Guessing you missed something. >>> >>> https://github.com/weewx/weewx/wiki/Cleaning-up-old-bad-data has the >>> wiki answer, but I do it a 'little' differently. Capturing it here for >>> future-me..... >>> >>> Here's how to query your tables to look for records you missed cleaning >>> up, using my db as an example. In my case I had a failing anemometer that >>> used to record absurdly high readings. Looks like I neglected to clean up >>> windGust back years ago when I cleaned up the sustained wind values, so I >>> have some crazy readings still in there for windGust that I'll use as an >>> example. >>> >>> # find daily maximums that were over a reasonable high for this location >>> pi@pi4:~/weewx-data/archive$ echo "select >>> datetime(dateTime,'unixepoch','localtime'),datetime,max from >>> archive_day_windGust where max > 50;" | sqlite3 vp2.sdb >>> 2018-03-27 00:00:00|1522134000|57.0 >>> 2018-04-15 00:00:00|1523775600|67.0 >>> 2018-04-17 00:00:00|1523948400|59.0 >>> 2018-09-11 00:00:00|1536649200|64.0 >>> 2018-09-22 00:00:00|1537599600|52.0 >>> >>> # verify those maximums came from archive table readings for those days >>> pi@pi4:~/weewx-data/archive$ echo "select >>> datetime(dateTime,'unixepoch','localtime'),datetime,windGust from archive >>> where windGust > 50;" | sqlite3 vp2.sdb >>> 2018-03-27 11:15:00|1522174500|55.0 >>> 2018-03-27 11:20:00|1522174800|55.0 >>> 2018-03-27 12:15:00|1522178100|57.0 >>> 2018-03-27 12:20:00|1522178400|57.0 >>> 2018-04-15 06:45:00|1523799900|67.0 >>> 2018-04-15 06:50:00|1523800200|67.0 >>> 2018-04-17 02:30:00|1523957400|59.0 >>> 2018-04-17 02:35:00|1523957700|59.0 >>> 2018-04-17 03:05:00|1523959500|52.0 >>> 2018-04-17 03:10:00|1523959800|52.0 >>> 2018-09-11 10:00:00|1536685200|53.0 >>> 2018-09-11 10:30:00|1536687000|54.0 >>> 2018-09-11 10:35:00|1536687300|64.0 >>> 2018-09-22 17:15:00|1537661700|52.0 >>> >>> # clean up the archive table - perhaps should set windGust=0 here in >>> this example >>> echo "update archive set windGust=NULL where windGust > 42;" | sqlite3 >>> vp2.sdb >>> >>> # rebuild-daily for just the 5 days in the output above..... >>> weectl database rebuild-daily --date=2018-03-27 >>> (and so on) >>> >>> Note - you should really work off a 'copy' of your db anytime you want >>> to modify it. It's a little interesting to me that weectl database doesn't >>> have an easy to use "--from filename.sdb" kind of option to pick the >>> filename of the 'copy' of the db that you want to mess with. I suppose it >>> can be faked with a bogus weewx config file to run against via the --config >>> option. Maybe others have better ease of use ideas. >>> >>> >>> On Saturday, June 7, 2025 at 7:07:22 AM UTC-7 RonV42 wrote: >>> >>>> I had to replace my Tempest station and the new one they sent didn't >>>> have the radiation and lux sensor properly calibrated. This effected the >>>> records in the Belchertown skin where I now have the radiation exposure of >>>> the planet Mercury as a record on May 27th. I set all the data values to >>>> null and rebuilt daily and fixed the graphs but the record for that >>>> measurement doesn't seem to want to clear and be re-canulated. Any idea's >>>> on how to fix this? >>>> >>>> Thanks >>>> >>> -- 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 visit https://groups.google.com/d/msgid/weewx-user/5a3c607f-d8af-43f0-80a1-541e84071804n%40googlegroups.com.
