Does Cumulus average / verify readings to stop spikes

[mattsimis]

Was reading how a Developer on FreeBSD made software to read the FineOffset and noticed it occasionally spikes to extremes (seems to be a common problem). He set his software to only update the data log if the info from the FO was verified at least twice in a row. Does Cumulus do this (I think not as I get spike occasionally) or will it at some stage?

[steve]

It just discards obviously invalid data. There's also an option to specify a maximum acceptable change between temperature readings - see 'EWTempdiff' in the cumulus.ini page in the wiki.

I may add further/more intelligent garbage checking in Cumulus 2.

[seh1]

I'd really like to see some additional data checks. I've got a Tycon rebranded FineOffset station and it occasionally generates bogus data, usually when it reconnects after a dropping the link between the station and receiver. Hand munging the data files is a bit cumbersome, particularly when I don't catch it for a day or two.

I just got back from a week's vacation and found that according to the weather station we'd had a heat wave of 999'F coupled with a 131 inches of rain. Glad I missed that one.

The station works well about 99% of the time, but that <1% is getting annoying.