Gene and all,
Having set up your solution, I’m now having a problem getting my Rachio controller to “stay” with my pws. The controller keeps reverting to another local weather station instead. Rachio says this could be due to my pws not responding or having null data.
Here’s my pws link:
https://www.pwsweather.com/obs/KSCMOUNT85.html
It does have some data. But it also says in red letters "there is no data for the period you requested."
Any clues what’s going on here?
@knwpsk it seems that you forgot to setup a cron-job to continuously transfer the data. The wufyi.com URL you’ve created on the third should be used at cron-job.org to be executed every 5 minutes or so. This will allow a data point to be transferred every time your wufyi.com custom URL is requested.
When setting up the cron-job, it is a good idea to enable failure notifications and the option to save results.
Feel free to contact me via support@wufyi.com in case you run into any further issues.
Cheers,
Gene
Yeah I overlooked that, assumed wufyi was running automagically.
I set it up now. Thanks!
2 Likes
The magic is that wufyi.com does not require you to store any data on the wufyi server. This is done this way in order to avoid terms of service issues with weather underground. Fingers crossed the site will be able to stay up and running for a good long while. 
Cheers,
Gene
4 Likes
It’s been a few weeks since I set up cron-job/wuyfi and I’m very pleased. Occasionally cron-job misses a report and occasionally I get something like this (which may be rounding since I’m in mm):
Here’s some WU data today from my Davis station, I’ve highlighted some rain data:
Here is the same data on PWS as reported by cron-job with the same rain data highlighted (ignore the timestamp, PWS have acknowledgrd a bug that cannot handle DST so reports everything an hour different during out summertime):
Weird eh? Any thoughts @franz?
Going forward a bit, here’s the data set again. The Wunderground rain data coincides with what appears on the local console, the PWS rain data looks screwy!
WU:
PWS:
Sorry to see that you having such an issue. I’ve known about occasional false data showing up in the transfer from weather underground, more on this here (link), but daily rain readings showing up 10 times smaller than they should is something new. I’ll try to investigate it, and let you know what I find.
Edit: It is definitely an issue with WU, here is what I’m getting via API (link), note the time of the data point is at 3:20PM, but it still reporting false rain data (just 0.02in / 1mm).
I’ll try to expedite one of the fixes I was considering, stay tuned.
Gene
1 Like
Thanks Gene. I saw on another thread that Rachio uses Solar Energy in its ET calcs for Flex.I see that the API has the Solar Energy data point but also a UV point. Two questions: Where does the SE point go (I don’t see it on the PWS page) and where does the UV point come from (My station measures solar but not UV - I baulked at the price of Davis’ UV sensor!)?
PS is there any significance on the letter colouring on the API output?
wufyi does transfer the solar / UV data over to pwsweather, but you are right, it is not displayed on their interface.
As far as I know @franz can’t get access to data as well, he talks about it here (link). Maybe that has since changed.
UV is an “index” calculated by WU, it is their guess how likely you are to get a sunburn 
I’ve seen formulas around the net for that calculation, not sure if they will match WU exactly.
Wufyi will support CWOP stations shortly so your solar data will be used sooner, rather then later.
As far as coloring on API output, it is done by jsonblob.com for easier differential between strings (green) and numerical values (red).
Gene
1 Like
It’s never easy…Rachio can’t get solar data from WU via PWS but can via CWOP/APRS. However from CWOP Rachio uses rain in the last 24 hours field not rain since last midnight and my Davis pws reports the 24 hour figure 5 times larger than actual (but the figure since midnight is spot on) and Davis haven’t worked out why yet…Gaaaah!
Got it up and running with my Acurite 5n1! Thanks for the instructions!
1 Like
Next instalment. I was looking at the wufyi’d WU data as it appears on PWS for the last couple of hours last night. Here’s the screenshot:
The odd things going on here (besides PWS showing times for GMT rather than GMT+1 which is local time for the station) are:
- Rain today (last column) is accurate up to 11.05pm (which is 00.05 12 Jul in local time) showing 16.5mm for the day (allowing for a few minutes lag in the system).
- Rachio picked up the total rain as 13.5mm which is the figure for 10 before midnight (circled in orange) Remember although the Rachio app screenshot says 1.35cm for the 10th that’s a bug and is really the 11th…
- There is an odd hiatus for an hour when 103.4mm appears (green circle) and then the number reverts at 11.30pm (00.30 local time) to the same number carried by the local weather station console and remains fine for the rest of 12 July (so far)
Actually they are very helpful. I’m fine with people knowing the issues wufyi has to deal with, WU api is far from perfect.
Original post deleted. Summary:
I’ve attempted to fix issues with WU api by using a rapid fire data source. A testing area was setup at wufyi,com/v2/, but since than a serious limitation has been discovered. In order to avoid having WU block some of the user functionality, I’m removing the test code for now. v2 will temporarily use older, tested, code. Until the nature of the wu block is fully understood.
Thank you for understanding,
Gene
Sorry @Gene I’ve been out and about and only just seen this. Have just switched over the cron-job to v2 and will see how it goes. Not expecting rain today so it may take a little while before we see how the rain posts have been affected (you know, it’s summer here in the UK - near desert conditions, hardly ever rains…!)
I’ve only had two transfers through so far and seem to be getting a wacky dew point (5537.2C!!) other data points look OK. I’ll leave it on v2 for a while on the assumption that the rain is fixed.
@WokingWounded No worries, actually you may want to hold off on switching. I was just coming here to update the post. It seems that WU has some sort of limit in place for the interface that v2 is using. I woke up to my updater failing…
Please switch back to regular for now. I need more testing.
Edit: I’m going to try a different approach. Stay tuned.
@Gene. OK. Once again I’ve been away but I’ve just switched back. The v2 seems to have been working fine in the meantime - even the dew point readings settled down to something sensible after an hour or so. Only odd thing is that normally cron-job sends a message if a transfer fails. I’ve had none in the last 24 hours but inspecting the PWS data there is a mixture of 5 and 10 minute gaps between readings (see for instance 15 June 03.40am GMT 05.30am GMT etc) even though cron-job is set to 5 minute intervals.
Hi buddy,
Just FYI, I’m still testing two new approaches. I’ll see which one deserves a further study.
So far V2 provided unreliable, but not completely useless. I’ll see if I can integrate it as a double-check scenario
Test data available here: https://www.pwsweather.com/obs/WUSTATION.html
it is based on https://www.wunderground.com/personal-weather-station/dashboard?ID=KFLBOCAR130
I’m also testing V3 which uses a different method, also independent of API
Test data available here: https://www.pwsweather.com/obs/TSTSTATION.html
it is based on https://www.wunderground.com/personal-weather-station/dashboard?ID=KFLFORTP31
If any of these provide promising, I’ll send you an invite. Sorry I’m keeping address secret for now (/v2/ points to old code), but as you can see I’m testing several options.
Cheers,
Gene
3 Likes