meteohub.de

First, thanks for adding the humidity adjustment. My humidity levels are off, and the adjustment provides me a means to provide more accurate humidity and dew point numbers.

I realize that adjustments to humidity are different that temp, BP, etc, as humidity has an absolute minimum of 0% and maximum of 100%. Your solution is to provide the endpoints that will be set to 0 or 100, which will keep the numbers in rage (which is good), but you also scale the in-between numbers, which is causing problems for me.

Like my temperature and BP adjustments, my humidity numbers always need to be adjusted up/down the same amount (not scaled). The scaling is causing my numbers to look great when the humidity level is low, not as good when in the middle, and not good at all when they are high. This is because I get the full adjustment when the humidity is low, about half the adjustment when the humidity sensor is at 50%, and not much at all when the sensor is high. With the current adjustment setting I don’t think I can get the same adjustment throughout the entire range.

I realize that because humidity has the hard 0 and 100 endpoints the adjustment to humidity can’t be as clean as temperature or BP, but could the following be performed instead of the scaling (which doesn’t work properly for numbers as they scale away from the endpoints):

Provide a singe field for humidity adjustment in +/- n.n% If the sensor reads less than .5% then report 0%, done. If the sensor reads 99.5% or higher then report 100%, done. Otherwise, adjust the sensor reading based on the supplied adjustment value. If the adjusted value is less than .5% the report 0%, done. If the adjusted value is 99.5% or higher then report 100%, done. Otherwise report the adjusted value.

This would ensure the reported value never goes below 0 or above 100%, and the fixed adjustment value is used as much as possible. I do realize at the endpoints there will be a jump up/down in the humidity to 0 or 100%, but that can’t be helped. The real benefit is that the adjustment value is used in the calculation for most of the entire range of humidity values.

Again, I know using an algorithm like I indicated above would fix my problems (where my low humidity values are adjusted properly but as the humidity increases the adjustment gets worse and worse), and the procedure works closer to the way temperature and BP adjustments are made.

As you can set adjustment values below 0 and above 100 your needs can be solved by the current procedure. When you want to add 5% on any value, then just specify -5 and 95 as the bounds to be mapped to 0 and 100.

Giving factor and offset in a f(x) = a*x + b fashion is something many users can't easily deal with, but the given method can express exactly the same linear equation.

OK, I think I see what you are doing. In my case I wanted to reduce the humidity level by 6, so I had 6 and 100 as the bounds to be mapped to 0 and 100, so based on you information I changed it to 6 and 106 and it appears to be working properly. I checked my reported vs. analysis numbers and they are looking much better - thanks.

I still have one question. The sensor will supply data that always falls in the range of 0 to 100, so with my settings, when it’s raining out and there’s a humidity level of 100%, will the reported value be 94% humidity? If so, can you make a minor adjustment to your algorithm where if the sensor value is less than .5% you report 0 and if the sensor value is greater than or equal to 99.5% you report 100? Without that adjustment it makes the reported value appear very odd (and makes it looks like the humidity sensor is broken). I’m pretty sure that the AcuRite algorithm doesn’t have this last adjustment so when their customers make an adjustment and it’s raining out they report that their humidity sensor is broken because it never reports 100% humidity.

What do you think? Again, thanks for explaining how to get the intermediate numbers correct.