NOAA Plugin failing

[jay (support)]

@dtich, KBUR hasn't updated since late last night - it's definitely down for some reason.

[dtich]

i see this. and of course no one is answering at the local nws/noaa office. ever. so i have no idea the issue and when it might be resolved for good.

so. i should switch to another station. kvny is similar in forecast and conditions usually. not very far.

can you help me with the python syntax for the action to insert variable into indigo var - at the moment i do it with insert device state into var action pulldown, but if i'm going to switch stations back and forth i'd rather do a script and possibly automate it, like if no update in values here, get values from other station temporarily kinda thing. i'll work on that, but if you can start me with the syntax for, say:

insert state (temperaturef) for device "weather station_kbur" into variable: weathercurrenttemp

starting to cobble together in babytalkpy. thoughts appreciated

Code: Select all

##var ref
#indigo.devices[204052749] # "weather station_van nuys"
#indigo.devices[228772392] # "weather station_burbank"
#indigo.variables[41620001] # "Weather_Condition"
#indigo.variables[162814321] # "WeatherCurrentHumidity"
#indigo.variables[1354783145] # "WeatherCurrentTemp"
#indigo.variables[1787162613] # "WeatherCurrentWindSpeed"

##init stations
kbur = indigo.devices[228772392]
kvny = indigo.devices[204052749]

##determine which station to use-- if bur is updated more recently than vny, use bur, else use vny
#curdatetime #(this has to be local i think), also maybe don't want to use this anyway
if kbur.states['lastChanged'] > kvny.states['lastChanged']
        wthrStation = indigo.devices[228772392]
else:
        wthrStation = indigo.devices[204052749]

##set temp vars from station info       ###does this need str() ?
cnttemp = wthrStation.states['temperatureF']
cnthumid = wthrStation.states['humidity']
cntwind = wthrStation.states['windMPH']
cntcond= wthrStation.states['currentCondition']

##update ind vars
indigo.variable.updateValue(1354783145, str(cnttemp))
indigo.variable.updateValue(162814321, str(cnthumid))
indigo.variable.updateValue(1787162613, str(cntwind))
indigo.variable.updateValue(41620001, str(cntcond))

[jay (support)]

The variable updates in that script look fine. One small issue: lastChanged is a device property, not a state. Corrected script:

Code: Select all

##var ref
#indigo.devices[204052749] # "weather station_van nuys"
#indigo.devices[228772392] # "weather station_burbank"
#indigo.variables[41620001] # "Weather_Condition"
#indigo.variables[162814321] # "WeatherCurrentHumidity"
#indigo.variables[1354783145] # "WeatherCurrentTemp"
#indigo.variables[1787162613] # "WeatherCurrentWindSpeed"

##init stations
kbur = indigo.devices[228772392]
kvny = indigo.devices[204052749]

##determine which station to use-- if bur is updated more recently than vny, use bur, else use vny
#curdatetime #(this has to be local i think), also maybe don't want to use this anyway
if kbur.lastChanged > kvny.lastChanged:
    wthrStation = indigo.devices[228772392]
else:
    wthrStation = indigo.devices[204052749]

##set temp vars from station info       ###does this need str() ?
cnttemp = wthrStation.states['temperatureF']
cnthumid = wthrStation.states['humidity']
cntwind = wthrStation.states['windMPH']
cntcond= wthrStation.states['currentCondition']

##update ind vars
indigo.variable.updateValue(1354783145, str(cnttemp))
indigo.variable.updateValue(162814321, str(cnthumid))
indigo.variable.updateValue(1787162613, str(cntwind))
indigo.variable.updateValue(41620001, str(cntcond))

[dtich]

learning!

this is throwing syntax err:

Code: Select all

if kbur.lastChanged > kvny.lastChanged

should it do it with actual vars, not temps?

like:

if indigo.devices.lastChanged[228772392] > ... ?

or, hasattr()?


thx!

[dtich]

nope. it was colon missing derp

[dtich]

added few more lines to indicate selected station. now have the system select among two or three nearby stations whichever is latest update, then displays which one is selected for calculations (currently KVNY) and a nearby one for ref (currently KUSC). feel more secure than relying on one station.. esp given how bloody unreliable they seem to be now.

thanks for help jay.

[dtich]

Jay, can you show me a line that compares lastChanged to currentDateTime, like, if lastChanged is more than 60mins ago, or 120mins ago?

might want to do it that way instead at some point.

thx

looking at the dox, is it something like this?:

from datetime import datetime
timeDelta = datetime.now() - kbur.lastChanged
if timeDelta.minutes > 60:
******
EDIT: ok, so i'm here:

Code: Select all

from datetime import datetime
timeDelta = datetime.now() - kbur.lastChanged

if timeDelta.seconds > 7200:
#if kvny.lastChanged > kbur.lastChanged:
	wthrStation = indigo.devices[204052749]
	indigo.variable.updateValue(752974756, "KVNY") # "WeatherCurrentStation"
	indigo.server.log("KBUR Time Since Update: "+ str(timeDelta), type="Irrigation:")

gives me:

Jan 28, 2022 at 3:44:07 PM
Irrigation: KBUR Time Since Update: 15:14:00.498093

is that hours:mins:seconds?

i can prob live with that. but, in case i wanted to round the seconds, or just do hours:mins.. how i do that? thx!

[jay (support)]

Sure:

Code: Select all

from datetime import datetime
timeDelta = datetime.now() - kbur.lastChanged

if timeDelta.seconds > 7200:
#if kvny.lastChanged > kbur.lastChanged:
    wthrStation = indigo.devices[204052749]
    indigo.variable.updateValue(752974756, "KVNY") # "WeatherCurrentStation"
    indigo.server.log("KBUR Time Since Update: "+ str(timeDelta.strftime("%H:%M")), type="Irrigation:")

strftime is a cool function with lots of options.

[dtich]

ahhh yes. that is a handy function.

this:

Code: Select all

indigo.server.log("KBUR Time Since Update: "+ str(timeDelta.strftime("%H:%M")), type="Irrigation:")

is giving me err:

AttributeError: 'datetime.timedelta' object has no attribute 'strftime'

doesn't like my var. does it only work on an actual datetime return? i tried several other arrangements and some other vars, no joy.

[dtich]

Seems the code for calculating time since update is only within 24 hours, doesn't count more than one day? Currently KBUR is still not updating, but the routine has promoted it to current because the time call is not taking the number of days into account. Any ideas how to get it to include days... ?

This is current:

Code: Select all

timeDelta = datetime.now() - kbur.lastChanged

if timeDelta.seconds > 7200:
#if kvny.lastChanged > kbur.lastChanged:
	wthrStation = indigo.devices[204052749]
	indigo.variable.updateValue(752974756, "KVNY") # "WeatherCurrentStation"
	indigo.server.log("KBUR Time Since Update: "+ str(timeDelta), type="Irrigation:")
	#indigo.server.log("KBUR Time Since Update: "+ str(timeDelta.strftime("%H:%M")), type="Irrigation:")

[Update.]
I did switch back to the logic using simple lastChanged greater than and that has put KVNY back to current as it should be. It also logs this:

Irrigation: KBUR Time Since Update: 1 day, 0:47:19.431777

That does work for now, but ultimately would like to be able to keep it to switching when BUR is offline for two hours no matter how many days are also tacked onto that 2 hours LOL.. can't find any explanation for KBUR being down for more than 24 hours now... But it clearly is. Doesn't even show up on the list of current METARS at this point. Odd.

Thanks for any pointers, no rush. Also, the title of this thread is somewhat over-dramatic it seems to me, certainly in my case, so if my comments want to be moved, or the thread re-titled (NOAA Plugin Station Update Issues?)... makes sense to me. 2¢.

[scs]

logs sent a couple minutes ago . I sent the 26th and 27th. the 26th is where the logging just stops at 21:16 , and then you can see where I restarted the plugin on the 27th . ... seems to support the idea that it may have hung.

I have a little Fibaro FGK-10X that supports a temp sensor, so that's an option if the NOAA problem continues

I was considering calling the airport and pointing out that I don't feel I'm getting my money's worth for this NOAA weather service, especially considering all the money I'm paying for it

[jay (support)]

Seems the code for calculating time since update is only within 24 hours, doesn't count more than one day? Currently KBUR is still not updating, but the routine has promoted it to current because the time call is not taking the number of days into account. Any ideas how to get it to include days... ?

This is current:

Code: Select all

timeDelta = datetime.now() - kbur.lastChanged

if timeDelta.seconds > 7200:
#if kvny.lastChanged > kbur.lastChanged:
	wthrStation = indigo.devices[204052749]
	indigo.variable.updateValue(752974756, "KVNY") # "WeatherCurrentStation"
	indigo.server.log("KBUR Time Since Update: "+ str(timeDelta), type="Irrigation:")
	#indigo.server.log("KBUR Time Since Update: "+ str(timeDelta.strftime("%H:%M")), type="Irrigation:")

[Update.]
I did switch back to the logic using simple lastChanged greater than and that has put KVNY back to current as it should be. It also logs this:

Irrigation: KBUR Time Since Update: 1 day, 0:47:19.431777

That does work for now, but ultimately would like to be able to keep it to switching when BUR is offline for two hours no matter how many days are also tacked onto that 2 hours LOL.. can't find any explanation for KBUR being down for more than 24 hours now... But it clearly is. Doesn't even show up on the list of current METARS at this point. Odd.

Thanks for any pointers, no rush. Also, the title of this thread is somewhat over-dramatic it seems to me, certainly in my case, so if my comments want to be moved, or the thread re-titled (NOAA Plugin Station Update Issues?)... makes sense to me. 2¢.

timedelta objects have 3 parts to them: days, seconds, and microseconds. So, you probably want something like this:

Code: Select all

if timeDelta.seconds > 7200 or timeDelta.days > 0:

Which would take into account the 24 hour rollover.

[dtich]

yup, perfect, thx jay, much appreciated

[jay (support)]

logs sent a couple minutes ago . I sent the 26th and 27th. the 26th is where the logging just stops at 21:16 , and then you can see where I restarted the plugin on the 27th . ... seems to support the idea that it may have hung.

I have a little Fibaro FGK-10X that supports a temp sensor, so that's an option if the NOAA problem continues

I was considering calling the airport and pointing out that I don't feel I'm getting my money's worth for this NOAA weather service, especially considering all the money I'm paying for it

Replied to your email.

[dtich]

IGNORE THIS POST PLS. I'LL KEEP IT POSTED FOR MY PERSONAL SHAME
So, explain me this one...:

Jan 29, 2022 at 11:53:13 AM
Script address : KBUR
batteryLevel : None
buttonGroupCount : 0
configured : True
description : burbank airport
deviceTypeId : station
displayStateId : temperatureString
displayStateImageSel : TemperatureSensor
displayStateValRaw : 50.0 F (10.0 C)
displayStateValUi : 50.0 F (10.0 C)
enabled : True
energyAccumBaseTime : None
energyAccumTimeDelta : None
energyAccumTotal : None
energyCurLevel : None
errorState :
folderId : 443424329
globalProps : MetaProps : (dict)
com.perceptiveautomation.indigoplugin.NOAAWeather : (dict)
address : KBUR (string)
id : 228772392
lastChanged : 2022-01-28 00:30:07
lastSuccessfulComm : 2022-01-28 00:30:07
model : Weather Station
name : weather station_burbank
ownerProps : com.perceptiveautomation.indigoplugin.NOAAWeather : (dict)
address : KBUR (string)
pluginId : com.perceptiveautomation.indigoplugin.NOAAWeather
pluginProps : emptyDict : (dict)
protocol : Plugin
remoteDisplay : True
sharedProps : com.indigodomo.indigoserver : (dict)
states : States : (dict)
currentCondition : Fair (string)
currentConditionIcon : nskc (string)
dewPointC : -5.6 (string)
dewPointF : 21.9 (string)
dewPointString : 21.9 F (-5.6 C) (string)
heatIndexC : - data unavailable - (string)
heatIndexF : - data unavailable - (string)
heatIndexString : - data unavailable - (string)
humidity : 33 (string)
latitude : 34.19967 (string)
location : Burbank - Bob Hope Airport, CA (string)
longitude : -118.36538 (string)
observationDate : Thu, 27 Jan 2022 23:53:00 -0800 (string)
pressureInches : 30.22 (string)
pressureMillibars : 1022.4 (string)
temperatureC : 10.0 (string)
temperatureF : 50.0 (string)
temperatureString : 50.0 F (10.0 C) (string)
visibility : 10.00 (string)
windDegrees : 20 (string)
windDirection : North (string)
windKnots : 3 (string)
windMPH : 3.5 (string)
windString : North at 3.5 MPH (3 KT) (string)
subModel :
subType :
supportsAllLightsOnOff : False
supportsAllOff : False
supportsStatusRequest : False
version : None

Jan 29, 2022 at 11:53:40 AM
Action Group weather/update vars
Irrigation: KBUR Time Since Update: 1 day, 11:23:33.306442
Irrigation: KBUR Last Obs: Sat, 29 Jan 2022 10:51:00 -0800

This shows a device props print out followed by a script run that logs the time delta for last change, and also the last observation date from the props/states.. but, the state in the device dump shows two days ago as expected, and the log shows... about half hour ago...?? Where is this coming from?

Here's the script lines that log the obs string:

Code: Select all

cntobsdate = wthrStation.states['observationDate']

#timeDeltaObs = datetime.now() - str(cntobsdate)
indigo.server.log("KBUR Last Obs: "+ str(cntobsdate), type="Irrigation:")

Huh?

This is the latest and last METAR from this station according to NOAA. It's so old it does not show up on aggregated current obs tables.

KBUR 280753Z 02003KT 10SM CLR 10/M06 A3022 RMK AO2 SLP224 T01001056 402220056 $

It shows last obs at 0753Z on the 28th, which is almost midnight on the 27th here, as expected by the last comm data in the plugin. So.. what gives?

Thx!

OH. Such a blunder.

Sorry guys. I realized just now that my var is set to the current one in the script, which at this point would be VNY.. so that explains the issue. So sorry for the panic. LOL