How does Weewx collect data?

[Andrea Di Saverio]

How does Weewx collect data?

Let's take my setup as example.
I have a Davis Vantage Vue with a data logger configured to store data every 1 minute.
In my Weewx configuration, the `loop_request` param in the `[Vantage]` section is set to 3, and the `archive_interval` in the `[StdArchive]` section is set to 300.

In this case, if I am correct, the `archive_interval` value is ignored, since I have my own configured in my data logger.

But what about the loop mode? As stated in the guide, `WeeWX will then start monitoring live sensor data (also referrred to as 'LOOP' data)`: so the data in the data logger are ignored?
Currently, in my setup, Weewx is storing data in my database every 1 minute and for sure this value comes from the data logger, since I didn't setup it in weewx.conf.


So:
- is my weewx storing data every 1 minute, because such data are fetched from data logger (which is configured to store data every 1 minute)?
- if true, what's the point of loop mode? how is it used by weewx?
- if false, why is weewx storing data every 1 minute? how frequently data are read in loop mode?


Thanks

[Tom Keffer]

Let's clarify the options:

loop_request: determines the type of LOOP packets to get. If you set it to 3, you will get a mix of legacy type 1, and the newer type 2, LOOP packets. This is covered in the [Vantage] section of the User's Guide.

archive_interval: as you say, for stations with hardware-specified archive intervals, such as the Davis stations, this is ignored. Instead, WeeWX will use the value in your hardware (which seems to be 1 minute). WeeWX then uses this value to determine when to stop requesting LOOP packets and, instead, process archive records. If a station can generate archive records in hardware, as your Davis Vue can, then those records will be taken from the hardware, then processed and stored in the database. 

To answer your questions:

1. Yes. So, in your case, every 60 seconds WeeWX will stop requesting LOOP packets then request any new archive records from your logger.

2. loop_request has nothing to do with the archive records stored in the logger. Changing it is useful because type 2 packets include some information that is not available in either type 1 packets, nor the archive record. Let's take an example: observation type "altimeter", It does not appear in type 1 LOOP packets, nor in archive records. If you set loop_request=1, then WeeWX will have to calculate it in software, then put it in the accumulators. If you set loop_request=2, then WeeWX does not have to calculate it and, instead, simply puts the value that was in the LOOP packet into the accumulators. Either way, at the end of an archive interval, a value is extracted from the accumulator and put in the archive record. With loop_request=1, the value will be from pure software, with loop_request=2, it will be from the Davis hardware.

3. It's true, so this question is moot. However, I will add that the frequency of LOOP packets is hardwired in the Davis stations at 2.5 seconds. It cannot be changed.

Hope this helps.

-tk

On Sun, Jul 31, 2022 at 9:52 AM Andrea Di Saverio <disaveri...@gmail.com> wrote:

How does Weewx collect data?

Let's take my setup as example.

I have a Davis Vantage Vue with a data logger, and it is configured to store data every 1 minute.

In my Weewx configuration, the `loop_request` param in the `[Vantage]` section is set to 3, and the `archive_interval` in the `[StdArchive]` section is set to 300.

If I am correct the `archive_interval` value is ignored, since I have my own configured in my data logger.

But what about the loop mode? As the guide states `WeeWX will then start monitoring live sensor data (also referrred to as 'LOOP' data)`: so the data in the data logger are ignored?

Currently, in my setup, Weewx is storing data in my database every 1 minute and for sure this value comes from the data logger, since I didn't setup it in weewx.conf.

So:

- is my weewx storing data every 1 minute, because such data are fetched from data logger (which is configured to store data every 1 minute)?

- if true, what's the point of loop mode? how is it used by weewx?

- if false, why is weewx storing data every 1 minute? how frequent data are read in loop mode?

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 weewx-user+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/weewx-user/aee9888a-1e27-4284-830d-59e8182fa914n%40googlegroups.com.

[Andrea Di Saverio]

Thanks for your explanation, but still I didn't get a point:

values in my weewx database are simply read from the data logger, or are computed starting from loop packets (loop packet -> accumulator -> value computation at the end of period)?

(Or maybe both cases, as mixed solution: available values are read from data logger, with higher priority, and not available ones are computed starting from loop packets data).

thx

[Tom Keffer]

1. If you set option record_generation to 'software', all archive records are extracted from the accumulators. Nothing comes from the logger.

2. If you set it to 'hardware', then archive records will be downloaded from the logger. The accumulators are then examined to see if there are any additional types in them that are not in the archive record. If so, these extra types are extracted from the accumulators and put in the archive record. 

In both cases, how observation types are extracted from the accumulators, or whether they are extracted at all, is configurable. See the wiki article Accumulators.

WeeWX was designed to work with lots of different types of hardware and to be very configurable. Hence the many stages and options.

To view this discussion on the web visit https://groups.google.com/d/msgid/weewx-user/c11f09ba-6844-40e3-bdae-bb77a57c00d7n%40googlegroups.com.

[Andrea Di Saverio]

Amazing.

Thanks