Hello,
the answer below:

Le jeudi 5 janvier 2017 15:48:32 UTC+1, Tom Keffer a écrit :
>
> these should match:
>
> SELECT sum FROM archive_day_rain WHERE dateTime = 1480114800;
>
> for datetime 1479942000, I get "2.48599999492855"
 

>
> SELECT SUM(rain) FROM archive WHERE dateTime > 1480114800 AND dateTime <= 
> (1480114800 + 86400);
>
> for datetime 1479942000, I get "1.24299999746428"
 
 

> If they don't, then something got corrupted in your daily summaries. 
> Rebuild them using the utility wee_database 
> <http://weewx.com/docs/utilities.htm#wee_database_utility>:
>
> wee_database --drop-daily --backfill-daily
>
> I did it with to commands. To rebuild the database took almost 1 hour 
(almost 9 years of data). Now the 2 lines above gives the same 
result "1.24299999746428"
Now I will follow the results and If a day it will happen again, I know 
what to do. 
On another way, What was happen in the archive we don't know. I kept the 
database if necessary.

Thanks to all for the support. It was an interesting analysis.

rds
Lio


-tk
>
> On Thu, Jan 5, 2017 at 5:38 AM, Andrew Milner <[email protected] 
> <javascript:>> wrote:
>
>> Did you delete-daily before you backfilled-daily??
>>
>>
>>
>> On Thursday, 5 January 2017 15:34:34 UTC+2, Andrew Milner wrote:
>>
>>> I'm leaving this to those who know your specific console.  It seems as 
>>> though the rain is being counted and totalled in two places somehow - 
>>> perhaps within the driver and also when the archive record is created.  
>>> I'll leave it to Matthew and Tom and those more knowledgeable than me.  The 
>>> counting difference will certainly be being caused by the Nulls - which 
>>> won't be counted.  
>>>
>>> ps -A  I think is the command you are looking for
>>>
>>>
>>>
>>>
>>>
>>> On Thursday, 5 January 2017 14:53:30 UTC+2, [email protected] wrote:
>>>
>>>> The archives doesn't include the value 0 because with the 2.7 version, 
>>>> I add a minimum of limitation in weewx.conf to 0.001 mm due to rain error 
>>>> calculation. I also did the same thing with the wind.
>>>> In the xls file, all the empty lignes are fill as '*NULL*' in the 
>>>> recorded data and archives.
>>>>
>>>> Normally, I get only one instance of weewx. Weewx is running on a 
>>>> raspberry. I already reboot the system. Do you know how see all programs 
>>>> are running?
>>>>
>>>> Le jeudi 5 janvier 2017 13:01:55 UTC+1, Andrew Milner a écrit :
>>>>>
>>>>> Something else does not look quite right with your archive_day_rain 
>>>>> table.  I would have expected the min to have been 0 unless it really did 
>>>>> rain in every archive record (which it didn't according to your 
>>>>> spreadsheet.  My table always has a zero for min.
>>>>>
>>>>> Are you sure that you have only got one instance of weewx running?  It 
>>>>> is starting to look as though somewhere,somehow your archive records are 
>>>>> being counted and totalled twice.  Personally I have not seen or heard of 
>>>>> that issue before.
>>>>>
>>>>> Sorry - I will have to let Matthew and/or Tom come to your rescue!!
>>>>>
>>>>>
>>>>>
>>>>> On Thursday, 5 January 2017 13:33:47 UTC+2, Andrew Milner wrote:
>>>>>
>>>>>> Looking at my archive_day_rain table the count is usually = 288 which 
>>>>>> is the number of archive records per day.  I archive every 5 minutes (= 
>>>>>> 12/.hr = 12*24/day = 288 per day)  The sum = the total of the data from 
>>>>>> the 
>>>>>> archive records for that day which starts at dateTime column
>>>>>>
>>>>>>
>>>>>>
>>>>>> On Thursday, 5 January 2017 13:22:46 UTC+2, [email protected] wrote:
>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Le jeudi 5 janvier 2017 10:49:44 UTC+1, Andrew Milner a écrit :
>>>>>>>>
>>>>>>>> So you are saying that the sum of the archived data in the archive 
>>>>>>>> table for the calendar day does not equal the value in the sum column 
>>>>>>>> in 
>>>>>>>> the archive_day-rain table?
>>>>>>>>
>>>>>>> Yes. 
>>>>>>>
>>>>>>>>
>>>>>>>> Does the value in he count column of archive_day_rain contain the 
>>>>>>>> correct number of archive records for the day which have been totalled 
>>>>>>>> to 
>>>>>>>> form the sum value??
>>>>>>>>
>>>>>>>  No I get 14 then archive_day_rain gives 28. See in the xls files in 
>>>>>>> attached. 
>>>>>>> What is the link between the sum of rain in a day and the number of 
>>>>>>> rain recorded in this day ?
>>>>>>>
>>>>>>> rds
>>>>>>> Lio
>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> On Thursday, 5 January 2017 10:41:03 UTC+2, [email protected] 
>>>>>>>> wrote:
>>>>>>>>
>>>>>>>>> Hello,
>>>>>>>>> Ok for this part of configuration.
>>>>>>>>>
>>>>>>>>> And concerning the sum of of the rain for a day, have you seen my 
>>>>>>>>> remarks in my last post?
>>>>>>>>> in summary, the SQL file (weewx.sdb), is made of different tables. 
>>>>>>>>> One is all the stored data (called archive), and others which are 
>>>>>>>>> statistics calculated by weewx.
>>>>>>>>>
>>>>>>>>> When I calculate the rain fallen during a day from the stored 
>>>>>>>>> data, this result which is closed to the other local weather station, 
>>>>>>>>> is 
>>>>>>>>> the half of the value calculated by weewx which is stored in the 
>>>>>>>>> archive_day_rain table. From my point of view there is a mistake when 
>>>>>>>>> weewx 
>>>>>>>>> calculates the rain fallen during a day.
>>>>>>>>>
>>>>>>>>> rds
>>>>>>>>> lio
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Le jeudi 5 janvier 2017 04:46:38 UTC+1, Andrew Milner a écrit :
>>>>>>>>>>
>>>>>>>>>> If your database is metric then rain is in cm
>>>>>>>>>> If your database is metricwx then rain is mm
>>>>>>>>>>
>>>>>>>>>> Your NOAA report is, according to the title line above the table, 
>>>>>>>>>> in mm - but the units are determined by the contents of the template 
>>>>>>>>>> and so 
>>>>>>>>>> can easily be changed to other units by changing the template 
>>>>>>>>>> appropriately.
>>>>>>>>>>
>>>>>>>>>> http://weewx.com/docs/customizing.htm#units 
>>>>>>>>>> <http://www.google.com/url?q=http%3A%2F%2Fweewx.com%2Fdocs%2Fcustomizing.htm%23units&sa=D&sntz=1&usg=AFQjCNHIPWGJUzXyeLDRewYXGmH3AhHQZg>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> On Thursday, 5 January 2017 00:19:57 UTC+2, [email protected] 
>>>>>>>>>> wrote:
>>>>>>>>>>
>>>>>>>>>>> Hi,
>>>>>>>>>>>
>>>>>>>>>>> Le mercredi 4 janvier 2017 00:38:20 UTC+1, mwall a écrit :
>>>>>>>>>>>>
>>>>>>>>>>>> On Tuesday, January 3, 2017 at 5:47:04 PM UTC-5, lionel.sylvie 
>>>>>>>>>>>> wrote:
>>>>>>>>>>>>
>>>>>>>>>>>> This is the same corrections in the weewx.conf between versions.
>>>>>>>>>>>>> Where I can find the commits?
>>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>> the change to rain multiplier happened on 11 jun 2015:
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>> https://github.com/weewx/weewx/commit/0bb7db1edc806caebe68357bc315ccf8674eea52
>>>>>>>>>>>>
>>>>>>>>>>>> so it has been part of the ws23xx driver for some time now.
>>>>>>>>>>>>
>>>>>>>>>>>> the divide-by-ten happens only to data from the logger, so if 
>>>>>>>>>>>> you are using record_generation=hardware you will definitely see a 
>>>>>>>>>>>> difference, but if you are using record_generation=software you 
>>>>>>>>>>>> will see a 
>>>>>>>>>>>> difference only when weewx has to do catchup.
>>>>>>>>>>>>
>>>>>>>>>>>>  I compare all the versions of the WS23xx I used since 2.7, all 
>>>>>>>>>>> are similar concerning the formula. The x10 in the commit above 
>>>>>>>>>>> concern the 
>>>>>>>>>>> wind speed which is included after the 2.7 version.
>>>>>>>>>>>
>>>>>>>>>>> To come back to my original question, it's concern not the 
>>>>>>>>>>> generated data but the values shown in the NOAA files.
>>>>>>>>>>> In the file attached in this post NOAA of November, when we take 
>>>>>>>>>>> the rain value for the 24 November 2016, this is 24.9mm. When we 
>>>>>>>>>>> take the 
>>>>>>>>>>> value in the database (weewx.sdb) for the same date ("1479942000" : 
>>>>>>>>>>> Sat Nov 
>>>>>>>>>>> 24 2016 00:00:00 GMT+0100), the value is 2.486. Is it the same 
>>>>>>>>>>> units?
>>>>>>>>>>> From a report of an official station close to mine (4-5kms), the 
>>>>>>>>>>> cumul was 11.2mm.
>>>>>>>>>>>
>>>>>>>>>>> http://www.infoclimat.fr/observations-meteo/archives/24/novembre/2016/saint-martin-d-heres/00014.html
>>>>>>>>>>>
>>>>>>>>>>> When I do the sum of the rain falls during this same day (data 
>>>>>>>>>>> taken in the archive table of weewx.sdb), I get 1.243 which is the 
>>>>>>>>>>> half 
>>>>>>>>>>> value form the archive_day_rain. Is it normal ?
>>>>>>>>>>> When I sum all the rain data for November 2016 coming from the 
>>>>>>>>>>> archive (= 8.029) and compare to the sum of the archive_day_rain 
>>>>>>>>>>> for 
>>>>>>>>>>> November (=16.058) and the NOAA of November (=160.6mm), I get a 
>>>>>>>>>>> double 
>>>>>>>>>>> value for the archive_day_rain and NOAA. Is there any issue during 
>>>>>>>>>>> the 
>>>>>>>>>>> calculation of the archive  ?
>>>>>>>>>>>
>>>>>>>>>>> rds
>>>>>>>>>>> Lio
>>>>>>>>>>> PS : After these all analysis, the archive_day_rain is in cm and 
>>>>>>>>>>> the NOAA is in mm. True? then I get the explanation for the x10 
>>>>>>>>>>> between 
>>>>>>>>>>> values of archive_day_rain and NOAA.
>>>>>>>>>>>
>>>>>>>>>>> -- 
>> 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 [email protected] <javascript:>.
>> 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 [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to