search

NORCatUofC / rain

An open-source data science project about rainfall in Chicago
MIT License
6 stars 7 forks source link

What does the 's' mean in the NOAA precipitation data #17

Open sbeslow opened 7 years ago

sbeslow commented 7 years ago

What does the 's' mean in the HOURLYPrecip column of the NOAA data? We were previously dropping it, but that is leaving some phony values for sure (6 inches in 1 hour! no way!). Now, I'm setting it to 0. Not sure if that is correct either though.

kbrose commented 7 years ago

The 's' means "This data value failed on of NCDC's quality control tests."

If you go here, select a year/month, and then select "VIEW DATA", then the bottom of the PDF that pops up has some documentation.

I started some notes on what I found out about the dataset. You can see them here: https://github.com/kbrose/rain/blob/master/n-year/notes.md

sbeslow commented 7 years ago

There it is! Thanks, Kevin. So how should we handle the 's' values? Assuming they should be zeroed?

kbrose commented 7 years ago

I would think so, yeah. That's the safest thing to do, anyway...