Message from @soup dad

Discord ID: 521852604460826624


2018-12-11 00:52:17 UTC  

actually ancap roads would be nicer than state funded roads tbqpfh

2018-12-11 00:52:21 UTC  

Mountain mama

2018-12-11 00:52:23 UTC  

Aye.

2018-12-11 00:52:31 UTC  

Jim.

2018-12-11 00:52:38 UTC  

@Ruger it’s a joke how Libertarians hate taxes and taxes pay for roads

2018-12-11 00:52:49 UTC  

Privatize the DMV.

2018-12-11 00:52:53 UTC  

HaHAA

2018-12-11 00:52:57 UTC  

so they hate taxes not roads

2018-12-11 00:53:47 UTC  

This is the meme I’m referring to, @Ruger

https://cdn.discordapp.com/attachments/418492114679365643/521852061504241674/image0.jpg

2018-12-11 00:54:25 UTC  

when a website gives a graph but no source data in an excel sheet

2018-12-11 00:54:30 UTC  

😤

2018-12-11 00:55:09 UTC  

It doesn't? Rio @soup dad

2018-12-11 00:55:12 UTC  

Rip

2018-12-11 00:55:13 UTC  

Fucking annoying

2018-12-11 00:55:23 UTC  

That's gay af

2018-12-11 00:55:30 UTC  

i need 2000-2018, and more accurate then a general line

2018-12-11 00:55:41 UTC  

I might just have to fucking eye ball it

2018-12-11 00:55:47 UTC  

but thats an ugly fucking data plot

2018-12-11 00:55:52 UTC  

Eye ball it, soup.

2018-12-11 00:55:56 UTC  

I already have to make up fucking data

2018-12-11 00:56:00 UTC  

its getting pretty inaccurate

2018-12-11 00:56:01 UTC  

It's going to be but

2018-12-11 00:56:10 UTC  

If you're making up data may as well

2018-12-11 00:56:27 UTC  

Making up is an ugly word, more like guessing with knowledge

2018-12-11 00:56:37 UTC  

mostly just using the naive model or the holt model for longer gaps

2018-12-11 00:56:50 UTC  

Nothing wrong with an educated guess

2018-12-11 00:56:57 UTC  

well its for a regression

2018-12-11 00:57:11 UTC  

I suppose, it's still going to be shifty either way.

2018-12-11 00:58:06 UTC  

comparing RAM prices in MB to a few things

I had the first data sets, but it was very inconsistent in its data collection, so you had holes for a month or a few months (one year only had 5 points) , and so for one month holes I used naive model (Just copying the one before hand) and holt (Pretty much just guessing by using previous trend)

2018-12-11 00:58:20 UTC  

Its accurate enough for government work, and you can use it for some guess work

2018-12-11 00:58:42 UTC  

Ahhh, right I see what you mean now.

2018-12-11 00:58:48 UTC  

But I then turned it into Quarterly data, to make it more precise by keeping as much of the original data in it

2018-12-11 00:59:05 UTC  

I like messing with Mike graham

2018-12-11 00:59:17 UTC  

I've been fucking with him on twitter with 2 different accounts for 2 months

2018-12-11 00:59:17 UTC  

But as youd guess, most this data i find is not in quarters, so I have to make it the yearly number the same 4 times over, so its got a few linear issues that make my data... uglyish

2018-12-11 00:59:48 UTC  

Welp her I go getting kicked

2018-12-11 00:59:50 UTC  

dim got called a nobody by the has been

2018-12-11 01:00:16 UTC  

@Ruger always

2018-12-11 01:00:34 UTC  

Stupid bot

2018-12-11 01:00:36 UTC  

it's a meme doe