Spring / Summer 2018 and onwards software

[Penri]

Good morning Hamish,

I hope the exam went well for you and all the hard work pays off.

In your question above are you asking what I'd like to see out of the overall status file or from each sensor?

Hwyl

Penri

[TerryJC]

Hamish,

The format that I used for the temperature readings is perfect for importing into a spreadsheet or data graphing tool.

[hamishmb]

Hi,

Penri:

Thanks, it did go well, I think

Both, if you know what is desired.

For example (for each sensor):

Code: Select all

#<PROBENAME>,<TIME>,<TICK>,<LEVEL_IN_MM>,<STATUS>
G4:M0,2018-06-11 09:25:28.197973,201,500,OK

Is that the kind of thing you had in mind?

Terry:

What format is that, and does it apply here because the values have different meanings? I must have missed that thread.

Hamish

[TerryJC]

Hamish,

It's in the Hardware Topic. It works because the first line acts as 'Table Headings' and so the data can be imported without any pre-processing.

[hamishmb]

Okay, that works. What I'm asking, really, is what values we want in each column - or do we just want as much as possible?

[TerryJC]

I'm just off shopping, so I can't answer in detail now. In principle we could have as many lines used for table headings as we like, if we need to group the results.

All we need to decide then is the order.

[hamishmb]

Okay.

As this point, I'm semi-ready to implement this, so I might just go ahead and we can adjust the order later then

[hamishmb]

Okay.

This is now sort-of implemented. The Reading class is now being used to represent readings as part of the implementation for this. I have made it so that the readings files are now in CSV format with this structure (taken from a test run at home):

Code: Select all

Start Time: 2018-06-11 12:08:49.594022

Starting to take readings...

TIME,SYSTEM TICK,ID,VALUE,STATUS

2018-06-11 12:08:49.651287,-1,SUMP:M0,0,OK
2018-06-11 12:08:56.652984,-1,G4:FS0,False,OK
2018-06-11 12:08:56.677747,-1,G4:M0,0,OK..
2018-06-11 12:09:11.694036,-1,G4:M0,0,OK..
2018-06-11 12:09:26.710709,-1,G4:M0,0,OK..
2018-06-11 12:09:41.727373,-1,G4:M0,0,OK..
2018-06-11 12:09:56.744013,-1,G4:M0,0,OK..
2018-06-11 12:10:11.761285,-1,G4:M0,0,OK..
2018-06-11 12:10:26.777392,-1,G4:M0,0,OK..
2018-06-11 12:10:41.794066,-1,G4:M0,0,OK..
2018-06-11 12:10:56.809557,-1,G4:M0,0,OK..
2018-06-11 12:11:11.826348,-1,G4:M0,0,OK

I can confirm that libreoffice calc can display these files nicely and I can easily construct graphs from the data. Some of the output duplication issues were solved as part of the implementation of this - though they aren't all fixed. I'm not sure this is ready for deployment tomorrow, but I will feel more sure of it after some extended tests later on in the week due to the number of changes I made. So I guess I could install the new software on Tuesday the 19th instead of tomorrow.

Note: I have not implemented the overall system CSV files, just the ones for each site - not each probe at this point.

Thoughts?

Hamish

[TerryJC]

Hamish,

I'd just finished manually producing an example with all the levels that I expect that we might have. It uses the two row headings principle to to provide only the Timestamp and the Levels / Flow Rate. I wonder if we should produce two logs; one like yours that provides a full status record and one like this which we can chart to spot big anomalies in level or flow rate.

See attached log file and spreadsheet, complete with a chart.

[TerryJC]

Drat...

The attachment bug strikes again.