I noticed this while setting up 3in1 and found that the temps in the detailed forecast (plaintext.txt) gave a temperature (highs and lows) that was different than is in the latest.txt / lastret.txt.
For instance the plaintext.txt file on Saturday states that the temp high for that day is 29, but the highest that it goes is 28.1 in the lastret.txt and latest.txt. For the low in plaintext.txt is says 13 and the latest.txt / lastret.txt both say 13.8.
Is this a rounding problem, plaintext.txt problem, or have I got something set wrong. I include the 3 files from one run for you to look at.
I have not checked, but I wonder if there may be other discrepancies too.
Good question, and one I answered extensively on another thread recently, which after a bried search I have failed to find! Can anybody remember where I posted that?
The very short answer is that it’s deliberate and appropriate, as it carefully takes into account short term temperature fluctuations (as a function of clouds, wind, etc.). Please use the value in the plaintext file as the actual high and low temperatures, NOT the highest and lowest number you’ve seen in the scrolling, hour-by-hour text.
Oh, this raises a question … I’ve been providing only the whole number values for highs and lows, as a “significant digits” issue (I can’t claim accuracy better than that - if that good). The decimal versions might be of interest to some users, though. Would anyone like these explicitly provided somewhere?