Re: [weewx-user] filling of gaps in measurements

[Andrew Milner]

I think this is all just to do with your modifications to simulator.  The 
simulator driver is not expected to ever handle real current data - the 
data is always generated within simulator.  If simulator is in simulation 
mode it generates loop data at 'normal' time intervals and if in generation 
mode as fast as it can.  Since most weather stations have a logging device 
I suspect the simulator driver emulates a catch-up process when resuming 
and generates data for the missing loop records from the last archive entry 
up to the current time.  I guess you need to stop this behaviour in your 
modifications!!  I guess you could just set the start time in weewx.conf 
simulator section to the current time and that could well force the issue.  
No doubt Tom knows a more subtle way to prevent a resumption/catch-up and 
force a start from current time.  The main problem is the mixing or real 
data with artificially generated data in the one 'driver'.  It may have 
been better to have used a different driver as the basis of your 
experiments/tests rather than simulator!!





On Thursday, 1 March 2018 17:05:25 UTC+2, Roebert Akraks wrote:

> Yesterday the only modification I did to the simulator.py is, that I 
> changed the genLoopPackets function to get the values from my station. This 
> worked in the evening, and I shut everything down and powered off the RPi. 
> Today in the morning, I started it up, and opened the weewx site with a 
> browser. The timestamp of the site was from yesterday in the evening, but 
> the data was already new (I noticed, because the barometer increased a lot 
> over night, and the temp was low as the sensor was outside in the morning 
> before I took it to the office). I pressed reload, and saw that the report 
> was now 5 minutes later then the previous one. Then I checked the syslog, 
> and saw, that every few seconds, a new report was generated (each time 5 
> minutes younger), until finally it had created all reports until today, 
> actual time. The graphs during this period have no gaps.
>
> http://roebert.selfhost.eu/zeuch/daytempdew.png
>
> I beleve this behaviour is because the simulator fills the gaps because 
> "start_time" and "resume_time" was specified somewhere. Normally the 
> simulator calculates a value for that given time and yields a package for 
> that given time, and therefore fills the gaps, but I changed it to use 
> actual data from the station.
> Today I changed that section to always use the current time, so if the 
> station is down, there are gaps. My question still is, where is that 
> start_time coming from? Is there a better way to remove this function then 
> just ignore it in the driver?
>
>
> Am Donnerstag, 1. März 2018 14:34:31 UTC+1 schrieb Andrew Milner:
>>
>> I don't fully follow the question either - but is the problem that you 
>> are saying the database is filled with current values for the period when 
>> the station was turned off - or that the plots have 'joined the dots' to 
>> make what appear to be continuous lines when you are expecting there to be 
>> gaps in the plots??
>>
>>
>>
>> On Thursday, 1 March 2018 15:09:24 UTC+2, Tom Keffer wrote:
>>
>>> I'm not completely following your question, but it sounds like this is 
>>> strictly an issue with the simulator. WeeWX itself is always in "realtime" 
>>> mode.
>>>
>>> If 'resume' is true, then the generator should continue where it left 
>>> off --- with the last timestamp in the database. It should never use the 
>>> "current values." If it seems like it is, that could be because the 
>>> simulation period is very long, so the values do not change very quickly.
>>>
>>> That's all I can offer without more details.
>>>
>>> -tk
>>>
>>>
>>> On Thu, Mar 1, 2018 at 1:18 AM, Roebert Akraks  wrote:
>>>
 Hi everyone,

 short introduction (skip if not interested):  :)
 I'm new to this group, my Name is Robert, I'm from Austria, and as I 
 work in a company producing (besides other things) temperature measurement 
 equipment, I'm building a weatherstation myself. Main reason is, that I 
 can 
 measure more acurate and I have the possibility to calibrate my equipment, 
 and the second reason is, that I own a WMR928, which is getting old and 
 has 
 to be replaced. 3rd, I want to do all the calculations myself, as some of 
 the methods or the Austrian institute of meteroloty (zamg.ac.at) uses 
 some other methods than used in most stations, and I want to calculate the 
 same way like the official stations here.
 The hardware is working so far, on the electronics side, I use an 
 Arduino, communicating via RS485 Modbus RTU with a RPi running weewx.
 My main problem: I'm not experienced with Python at all. With, what I 
 call "experimental programming" I managed to modify the simulator.py 
 driver 
 to read the values from the station, so basically everything is working 
 already.


 

Re: [weewx-user] filling of gaps in measurements

[Roebert Akraks]

Yesterday the only modification I did to the simulator.py is, that I 
changed the genLoopPackets function to get the values from my station. This 
worked in the evening, and I shut everything down and powered off the RPi. 
Today in the morning, I started it up, and opened the weewx site with a 
browser. The timestamp of the site was from yesterday in the evening, but 
the data was already new (I noticed, because the barometer increased a lot 
over night, and the temp was low as the sensor was outside in the morning 
before I took it to the office). I pressed reload, and saw that the report 
was now 5 minutes later then the previous one. Then I checked the syslog, 
and saw, that every few seconds, a new report was generated (each time 5 
minutes younger), until finally it had created all reports until today, 
actual time. The graphs during this period have no gaps.

http://roebert.selfhost.eu/zeuch/daytempdew.png

I beleve this behaviour is because the simulator fills the gaps because 
"start_time" and "resume_time" was specified somewhere. Normally the 
simulator calculates a value for that given time and yields a package for 
that given time, and therefore fills the gaps, but I changed it to use 
actual data from the station.
Today I changed that section to always use the current time, so if the 
station is down, there are gaps. My question still is, where is that 
start_time coming from? Is there a better way to remove this function then 
just ignore it in the driver?


Am Donnerstag, 1. März 2018 14:34:31 UTC+1 schrieb Andrew Milner:
>
> I don't fully follow the question either - but is the problem that you are 
> saying the database is filled with current values for the period when the 
> station was turned off - or that the plots have 'joined the dots' to make 
> what appear to be continuous lines when you are expecting there to be gaps 
> in the plots??
>
>
>
> On Thursday, 1 March 2018 15:09:24 UTC+2, Tom Keffer wrote:
>
>> I'm not completely following your question, but it sounds like this is 
>> strictly an issue with the simulator. WeeWX itself is always in "realtime" 
>> mode.
>>
>> If 'resume' is true, then the generator should continue where it left off 
>> --- with the last timestamp in the database. It should never use the 
>> "current values." If it seems like it is, that could be because the 
>> simulation period is very long, so the values do not change very quickly.
>>
>> That's all I can offer without more details.
>>
>> -tk
>>
>>
>> On Thu, Mar 1, 2018 at 1:18 AM, Roebert Akraks  wrote:
>>
>>> Hi everyone,
>>>
>>> short introduction (skip if not interested):  :)
>>> I'm new to this group, my Name is Robert, I'm from Austria, and as I 
>>> work in a company producing (besides other things) temperature measurement 
>>> equipment, I'm building a weatherstation myself. Main reason is, that I can 
>>> measure more acurate and I have the possibility to calibrate my equipment, 
>>> and the second reason is, that I own a WMR928, which is getting old and has 
>>> to be replaced. 3rd, I want to do all the calculations myself, as some of 
>>> the methods or the Austrian institute of meteroloty (zamg.ac.at) uses 
>>> some other methods than used in most stations, and I want to calculate the 
>>> same way like the official stations here.
>>> The hardware is working so far, on the electronics side, I use an 
>>> Arduino, communicating via RS485 Modbus RTU with a RPi running weewx.
>>> My main problem: I'm not experienced with Python at all. With, what I 
>>> call "experimental programming" I managed to modify the simulator.py driver 
>>> to read the values from the station, so basically everything is working 
>>> already.
>>>
>>>
>>> Now to the question:
>>> I noticed, that weewx automatically fills the gaps with current values, 
>>> in case the recording was interrupted. Eg. I shut down the station 
>>> yesterday in the evening, and started it today in the morning. Now It is 
>>> generating all missing reports from the evening till now, but it's using 
>>> the current values, which is not correct. I noticed, that the loader 
>>> function of simulator can be specified a resume time. I changed this to 
>>> always use the current time, but is there somewhere a better place to tell 
>>> weewx that it is in realtime mode? 
>>> Please excuse if this is very supid, but python is still like chinese 
>>> for me :)
>>>
>>> -- 
>>> 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.
>>> For more options, visit https://groups.google.com/d/optout.
>>>
>>
>> 

-- 
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 

Re: [weewx-user] filling of gaps in measurements

[Andrew Milner]

I don't fully follow the question either - but is the problem that you are 
saying the database is filled with current values for the period when the 
station was turned off - or that the plots have 'joined the dots' to make 
what appear to be continuous lines when you are expecting there to be gaps 
in the plots??



On Thursday, 1 March 2018 15:09:24 UTC+2, Tom Keffer wrote:

> I'm not completely following your question, but it sounds like this is 
> strictly an issue with the simulator. WeeWX itself is always in "realtime" 
> mode.
>
> If 'resume' is true, then the generator should continue where it left off 
> --- with the last timestamp in the database. It should never use the 
> "current values." If it seems like it is, that could be because the 
> simulation period is very long, so the values do not change very quickly.
>
> That's all I can offer without more details.
>
> -tk
>
>
> On Thu, Mar 1, 2018 at 1:18 AM, Roebert Akraks  > wrote:
>
>> Hi everyone,
>>
>> short introduction (skip if not interested):  :)
>> I'm new to this group, my Name is Robert, I'm from Austria, and as I work 
>> in a company producing (besides other things) temperature measurement 
>> equipment, I'm building a weatherstation myself. Main reason is, that I can 
>> measure more acurate and I have the possibility to calibrate my equipment, 
>> and the second reason is, that I own a WMR928, which is getting old and has 
>> to be replaced. 3rd, I want to do all the calculations myself, as some of 
>> the methods or the Austrian institute of meteroloty (zamg.ac.at) uses 
>> some other methods than used in most stations, and I want to calculate the 
>> same way like the official stations here.
>> The hardware is working so far, on the electronics side, I use an 
>> Arduino, communicating via RS485 Modbus RTU with a RPi running weewx.
>> My main problem: I'm not experienced with Python at all. With, what I 
>> call "experimental programming" I managed to modify the simulator.py driver 
>> to read the values from the station, so basically everything is working 
>> already.
>>
>>
>> Now to the question:
>> I noticed, that weewx automatically fills the gaps with current values, 
>> in case the recording was interrupted. Eg. I shut down the station 
>> yesterday in the evening, and started it today in the morning. Now It is 
>> generating all missing reports from the evening till now, but it's using 
>> the current values, which is not correct. I noticed, that the loader 
>> function of simulator can be specified a resume time. I changed this to 
>> always use the current time, but is there somewhere a better place to tell 
>> weewx that it is in realtime mode? 
>> Please excuse if this is very supid, but python is still like chinese for 
>> me :)
>>
>> -- 
>> 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 .
>> For more options, visit https://groups.google.com/d/optout.
>>
>
>

-- 
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+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [weewx-user] filling of gaps in measurements

[Thomas Keffer]

I'm not completely following your question, but it sounds like this is
strictly an issue with the simulator. WeeWX itself is always in "realtime"
mode.

If 'resume' is true, then the generator should continue where it left off
--- with the last timestamp in the database. It should never use the
"current values." If it seems like it is, that could be because the
simulation period is very long, so the values do not change very quickly.

That's all I can offer without more details.

-tk


On Thu, Mar 1, 2018 at 1:18 AM, Roebert Akraks  wrote:

> Hi everyone,
>
> short introduction (skip if not interested):  :)
> I'm new to this group, my Name is Robert, I'm from Austria, and as I work
> in a company producing (besides other things) temperature measurement
> equipment, I'm building a weatherstation myself. Main reason is, that I can
> measure more acurate and I have the possibility to calibrate my equipment,
> and the second reason is, that I own a WMR928, which is getting old and has
> to be replaced. 3rd, I want to do all the calculations myself, as some of
> the methods or the Austrian institute of meteroloty (zamg.ac.at) uses
> some other methods than used in most stations, and I want to calculate the
> same way like the official stations here.
> The hardware is working so far, on the electronics side, I use an Arduino,
> communicating via RS485 Modbus RTU with a RPi running weewx.
> My main problem: I'm not experienced with Python at all. With, what I call
> "experimental programming" I managed to modify the simulator.py driver to
> read the values from the station, so basically everything is working
> already.
>
>
> Now to the question:
> I noticed, that weewx automatically fills the gaps with current values, in
> case the recording was interrupted. Eg. I shut down the station yesterday
> in the evening, and started it today in the morning. Now It is generating
> all missing reports from the evening till now, but it's using the current
> values, which is not correct. I noticed, that the loader function of
> simulator can be specified a resume time. I changed this to always use the
> current time, but is there somewhere a better place to tell weewx that it
> is in realtime mode?
> Please excuse if this is very supid, but python is still like chinese for
> me :)
>
> --
> 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+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
>

-- 
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+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[weewx-user] filling of gaps in measurements

[Roebert Akraks]

Hi everyone,

short introduction (skip if not interested):  :)
I'm new to this group, my Name is Robert, I'm from Austria, and as I work 
in a company producing (besides other things) temperature measurement 
equipment, I'm building a weatherstation myself. Main reason is, that I can 
measure more acurate and I have the possibility to calibrate my equipment, 
and the second reason is, that I own a WMR928, which is getting old and has 
to be replaced. 3rd, I want to do all the calculations myself, as some of 
the methods or the Austrian institute of meteroloty (zamg.ac.at) uses some 
other methods than used in most stations, and I want to calculate the same 
way like the official stations here.
The hardware is working so far, on the electronics side, I use an Arduino, 
communicating via RS485 Modbus RTU with a RPi running weewx.
My main problem: I'm not experienced with Python at all. With, what I call 
"experimental programming" I managed to modify the simulator.py driver to 
read the values from the station, so basically everything is working 
already.


Now to the question:
I noticed, that weewx automatically fills the gaps with current values, in 
case the recording was interrupted. Eg. I shut down the station yesterday 
in the evening, and started it today in the morning. Now It is generating 
all missing reports from the evening till now, but it's using the current 
values, which is not correct. I noticed, that the loader function of 
simulator can be specified a resume time. I changed this to always use the 
current time, but is there somewhere a better place to tell weewx that it 
is in realtime mode? 
Please excuse if this is very supid, but python is still like chinese for 
me :)

-- 
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+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.