search

CodeForFoco / solar-scorecard

Solar Scorecard project: track Fort Collins renewable/solar energy goals
http://solar-scorecard.org/
MIT License
4 stars 6 forks source link

Get some fake data #7

Closed abettermap closed 7 years ago

abettermap commented 7 years ago

Might not have real data for awhile, so let's make something realistic. Maybe using Faker.

abettermap commented 7 years ago

Actually, I don't think Faker would be that useful for us. We really just need a small handful of custom columns. Turns out d3 actually has a built-in thing we might be able to leverage: https://github.com/d3/d3/blob/master/API.md#random-numbers-d3-random

deidyomega commented 7 years ago

My last pull request (https://github.com/CodeForFoco/solar-scorecard/pull/15), I created a simple line_generator() function that does a pretty good job of creating randomish lines with a high R^2 (meaning it's a straightish line).

If you guys can tell me the approx values of the real data, I can have my line_generator() create numbers within 1 - 2 stdevs of that approximation.

abettermap commented 7 years ago

Cool, let's hit Ian up on Slack to see if he can produce some values.

ijskor commented 7 years ago

Solaredge Sample Data.xlsx

ijskor commented 7 years ago

Try some of that, I can get more if needed. We will be collecting the 'kW' data primarily.

deidyomega commented 7 years ago

I posted in the slack, but the sample data has massive changes in kw / month. Unless I'm reading this excel incorrectly.

If this spike it a trend that's expected to continue, I'll need to create a exponential regression. Or do you think that now the new stations are online a linear model will be a more accurate projection?