I do a bit of both, but mostly the old‑school way. I keep a color‑coded diary where each day’s pressure trend gets a mood tag, and I let the clouds do the heavy lifting. When the barometer drops faster than a 5‑minute storm, I jot it down. I’ve tried a simple linear regression on a few months of data, but it’s usually the quick look at cumulonimbus that tells me what’s coming. If the pressure keeps swinging, I’ll pull up my spreadsheet, run a quick script, and see if the trend matches a past storm. But I’m still patient with the clouds and not too fond of slow elevators—pressure spikes are the only thing that gets my heart racing.