# It’s an algorithm thing x 2

Posted on February 22, 2016

First, weather:   we have no “special weather statements” up.   Weekend temps were springlike – fifties!  Today 45!   Sunny and beautiful!!   The front page of “weather channel” has niceness today and tomorrow… tomorrow night:   Rain/Wind Late.

I know what I’d be thinking.   Might be wet Wednesday.  Find raingear. (Not umbrella on the windy prairie, thanks.)

I follow a weather nerd on Twitter who posted sometime Saturday about the flakes hitting the fan Tues or Wed… and now, *if* you look at the long term forecast, you can see that yes, we’re expecting 100% chance of snow, 3-5 inches, w/ 28 mph winds.

Now, the other weather-attuned teachers have postponed exams and prepped students for “if you can’t get here Wednesday,” but lots of them have no idea.

If the weather forecast were being done by humans instead of computer generated algorithms, there *would be* a “special weather statement.

—-

… and student brings in stats book for help.  Someone else helping had her switch off x and y variables in plotting a linear regression.  That student knew the formulas for the other terms that I had to tell her to look up… but … I know input from output.   If you’re a helper who has an algorithm for “key words” (had to do with “an important variable is…” meaning that was your “y”, somehow), it’s just not going to hold up all the time.   Since the data was designed to help you predict a bear’s weight from its approximate height (since, as the problem said, it’s a little hard to weigh bears in the wild)… I coudl state with confidence that the height should stay in the x column…