Spurious data being logged

[dataworx]

I've been using Weatherlink for quite a while, with my station's data being uploaded to Weather Underground. I've been having problems with Weatherlink recently and have long been wanting to try Cumulus, so I installed V 1.9.4.
Everything works well, data gets uploaded to Wunderground as expected, but I have an error with rainfall figures reported to and graphed on Wunderground. We are subjected to frequent, scheduled power outages (3 per day, of 2 hours each.....) so I close Cumulus and power my PC down just prior to power being cut. When mains power is restored, I power the PC on, start Cumulus and verify that data is being downloaded from the logger. For some reason, spurious rainfall data is introduced, gets sent on to Wunderground and is posted online.

This behaviour (spurious data) occurs whenever I close Cumulus for a period of time - as a test, I earlier closed Cumulus for a while, kept the PC running and then opened
Cumulus again. Logged data was downloaded as expected, but included the odd rainfall figure of 127mm - see below:

13 12 19,11:35,21.6,73,16.6,6.4,11.3,144,0.0,127.0,1010.70,439.7,24.6,58,11.3,21.6,21.6,0.0,0,0.00,49.02,22.6,1064,0.0,135,0.0,127.0
13 12 19,11:40,21.9,70,16.2,8.0,14.5,135,0.0,127.0,1010.70,439.7,24.7,58,14.5,21.9,21.9,0.0,0,0.00,49.02,22.4,1070,0.0,135,0.0,127.0

I also get error messages such as this on Cumulus but I have no idea how to deal with these:

Error on line 1 of dayfile.txt: '''' is not a valid integer value
12 Dec 10:22:46 PM : Error plotting charts at line 1 of dayfile.txt: '''' is not a valid integer value
12 Dec 10:22:46 PM : Please correct the error by editing dayfile.txt

This results in the graph as shown below being published on Wunderground - the first event is when Cumulus was off earlier due to power failure, the second event is the test that I initiated:

RainCapture.PNG

I understand that the best solution may be to keep Cumulus running full time, but this is not an option at the moment. Is there any other action I can take to remedy this problem?

[mcrossley]

You need to fix your dayfile.txt file, it has some sort of corruption in it - at line 1

[dataworx]

Here is the first line of the file found at C:Cumulus/data/dayfile. I've had a good look at that file but cannot work out what the problem is:
10 12 19,40.2,0,00:01,16.4,23:51,19.2,21:08,1011.35,22:15,1011.62,22:45,0.0,00:00,0.0,17.9,32.8,16.1,21:52,76,21:14,86,23:53,0.00,0.0,19.2,21:08,18.7,21:08,16.1,23:29,0.0,00:00,14.9,23:34,15.3,21:52,13.9,23:51,116,0.1,0.0,0,00:00,0.0,00:00

The error report says: '''' is not a valid integer value. since neither the file nor line 1 contain these characters, I'm really not sure what I should be looking for. An integer is defined as a whole number so that is also not a good clue

[mcrossley]

So is your date format "10 12 19" ?
Please post your MXdiags folder, and your dayfile

[dataworx]

Yes, date format is 10 12 19

Here are the full contents of dayfile:

10 12 19,40.2,0,00:01,16.4,23:51,19.2,21:08,1011.35,22:15,1011.62,22:45,0.0,00:00,0.0,17.9,32.8,16.1,21:52,76,21:14,86,23:53,0.00,0.0,19.2,21:08,18.7,21:08,16.1,23:29,0.0,00:00,14.9,23:34,15.3,21:52,13.9,23:51,116,0.1,0.0,0,00:00,0.0,00:00
11 12 19,32.2,118,19:18,15.7,04:54,21.3,14:20,1010.60,17:30,1013.48,22:14,0.0,00:00,0.0,18.8,156.6,20.9,12:00,77,07:58,90,05:05,0.00,0.0,21.3,14:20,22.9,14:20,16.9,04:54,0.0,00:00,15.7,04:54,17.3,14:20,13.7,04:49,119,0.4,0.9,0,00:00,0.0,00:00
12 12 19,35.4,135,16:15,16.9,07:33,23.4,11:29,1010.40,16:14,1014.26,08:14,0.0,00:00,0.0,19.1,261.9,27.4,16:15,64,11:29,87,01:44,0.00,0.0,23.4,11:29,22.7,11:54,16.0,21:40,0.0,00:00,15.0,22:32,16.6,11:28,14.3,05:14,121,0.3,1.1,0,00:00,0.0,00:00

In C:Cumulus/Diags I find files listed cumulus.001 through to cumulus.009. Each file,when opened in a text editor, is over 7000 lines, so I'm reluctant to try paste it here. Can I attach the folder/file somehow? Edit: found the utility

[dataworx]

Hopefully this is what you need:

cumulus.001

dayfile.txt

Just to be clear, I'm using Cumulus 1, not MX

[mcrossley]

It is something to do with your odd unusual date formats, but as it is Cumulus 1, that is all I can say.

[dataworx]

Thanks for your time Mark.
Since I cannot change the date format on the Davis console, my only option is to change it in Windows' regional settings.
What is the common format when viewing a 'normal' dayfile?

[dataworx]

Seems to be all sorted now - it was a case of RTFM