Could this not be applied in a semi-similar method to calc_rainRate's
previous record look-up? E.g., psuedocoding,
Grab last archive value (T,Td,etc)
If abs(last archive value - current value)>some_value:
add suspect_flag
Then go back and remove suspect_flags based on whether a spike occurred?
archive_vals=[val for val in records[-n:-5]]
if any(archive_vales.suspect_flag):
inds=where(suspect_flag)
if abs(archive_vals[inds]-archive_vales[inds+1])<some_value:
suspect_flag=false
Obviously this requires a suspect_flag added to the sdb. I have no sql
experience, but this should be rather trivial I would think. Also, this
doesn't help issues with consecutive erroneous values, but adding that
should be trivial also.
On Sunday, June 23, 2019 at 4:14:49 PM UTC-5, Thomas Keffer wrote:
>
> Hello, Jared
>
> Unfortunately, StdQC can only reject values that are outside of a [min,
> max] range.
>
> However, this would not be a very hard service to write. Something like
> (NOT TESTED):
>
> import time
> from weewx.engine import StdService
>
> class MyQC(StdService):
> """Check for time and temperature out of bounds """
>
> def __init__(self, engine, config_dict):
> super(MyQC, self).__init__(engine, config_dict)
>
> self.bind(weewx.NEW_ARCHIVE_RECORD, self.new_archive_record)
>
> def new_archive_record(self, event):
> """Check for a high temperature at night"""
> # Make sure there
> if 'outTemp' in event.record and event.record['outTemp'] is not
> None:
> time_tuple = time.localtime(event.record['dateTime'])
> hour = time_tuple.tm_hour
> # This assumes US Customary units. You probably want to check
> the unit system first!!
> if (hour < 5 or hour > 20) and event.record['outTemp'] > 100:
> event.record['outTemp'] = None
>
> This example uses a pretty stupid time / temperature check. You'd probably
> want to substitute something more sophisticated.
>
> -tk
>
> On Sun, Jun 23, 2019 at 2:00 PM Jared Marquis <[email protected]
> <javascript:>> wrote:
>
>> I've noticed a few issues with the occasional bad temperature/dew point
>> measurement. Specifically, this week, my high temperature is: 101.0°F
>> at 01:42:00 AM (Monday). While unusual, not impossible. That said, it
>> happened overnight, and was surrounded by temperatures in the 50s. This is
>> obviously an errant reading.
>>
>> I am currently recording information every 5min. How can I look at a
>> time series of, say 5-10 temperature readings, to remove or flag suspect
>> data? Also, I'd like to be able to restrict tendencies to do the same
>> (e.g., +/-25* change in 5min is suspect and needs to examined with time
>> series when more data comes in).
>>
>> Is there any flag variable in the observation objects or would this need
>> to be added? Is it possible to examine other times with a qc function?
>>
>> 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] <javascript:>.
>> To view this discussion on the web visit
>> https://groups.google.com/d/msgid/weewx-user/b9af5ef9-66e7-41a6-8892-60f580f80b72%40googlegroups.com
>>
>> <https://groups.google.com/d/msgid/weewx-user/b9af5ef9-66e7-41a6-8892-60f580f80b72%40googlegroups.com?utm_medium=email&utm_source=footer>
>> .
>> For more options, visit https://groups.google.com/d/optout.
>>
>
--
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/16f4e521-b873-4703-b70e-9137b96a3ac4%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
