Filtering bad data in query functions

I have a typical TIG stack collecting SNMP data from some APC devices, which includes a few NetBotz wireless temperature sensors. I’m having an issue where if one of the wireless sensors loses signal it will report a temperature of -0.1C which skews the data any mathematical functions like Median or Mean. I was hoping there was a way that I could filter out any time there’s an entry of -0.1 but I haven’t found any method yet, or perhaps there’s a better way to write my query that wouldn’t be effected by one of these sensors disconnecting. Any advice would be appreciated.

Example:
1578964561000000000 POC Row D - Hot 28.7 10.168.110.200 datacollector01 netbotz-poc
1578964561000000000 POC Row D - Cold -0.1 10.168.110.200 datacollector01 netbotz-poc
1578964561000000000 POC Row C - Hot 28.3 10.168.110.200 datacollector01 netbotz-poc
1578964561000000000 POC Row C - Cold 20.3 10.168.110.200 datacollector01 netbotz-poc
1578964561000000000 POC ROW B - Hot 27.8 10.168.110.200 datacollector01 netbotz-poc
1578964561000000000 POC ROW B - Cold 22.6 10.168.110.200 datacollector01 netbotz-poc
1578964561000000000 POC ROW A - Hot 30.2 10.168.110.200 datacollector01 netbotz-poc
1578964561000000000 POC ROW A - Cold 23.1 10.168.110.200 datacollector01 netbotz-poc

SELECT mean(“Temperature”) FROM “Temperature Probes” WHERE (“Probe Name” =~ /Cold/) AND time >= now() - 3h GROUP BY time(10s) fill(null)

1578964530000000000 16.525 (should be closer to 22 without the bad data)

Did you try the not operator for Temperature?
Temperature != -0.1

I was able to work this in and it worked, thanks!

Glad to help you out.