Soares’s gists

Soares / Exploder.hs

Created April 10, 2012 14:46

How to explode GHC's brain

	{-# LANGUAGE ExistentialQuantification #-}
	data Test = forall a. Show a => Box a
	main = print a where Box a = undefined

Soares / vim.rb

Last active December 14, 2015 02:19 — forked from mgrouchy/vim.rb

	require 'formula'

	class Vim < Formula
	homepage 'http://www.vim.org/'
	url 'https://vim.googlecode.com/hg/', :revision => '6c318419e331'
	version '7.3.515'

	def features; %w(tiny small normal big huge) end
	def interp; %w(lua mzscheme perl python python3 tcl ruby) end

Soares / gist:10444320

Created April 11, 2014 06:34

Illiterate Botworld

	module Botworld where
	import Control.Applicative ((<$>), (<*>))
	import Control.Monad (join)
	import Control.Monad.Reader (Reader, asks)
	import Data.List (delete, elemIndices, intercalate, sortBy)
	import Data.List.Split (chunksOf)
	import Data.Maybe (catMaybes, isJust, fromMaybe, mapMaybe)
	import Data.Ord (comparing)
	import Text.Printf (printf)

Soares / keybase.md

Created March 17, 2015 02:38

Keybase proof

I hereby claim:

I am soares on github.
I am so8res (https://keybase.io/so8res) on keybase.
I have a public key whose fingerprint is D44F 9407 513A 18E4 6CDE E43F 6E2E 054D CFDD FBBB

To claim this, I am signing this object:

Soares / gist:a183d3bad69b533892b7

Created June 22, 2015 21:50

MacBook Pro WindowServer log messages after disconnecting external monitor (text is blurry afterwards)

	6/22/15 2:47:44.127 PM WindowServer[155]: Received display connect changed for display 0x4280a40
	6/22/15 2:47:44.128 PM WindowServer[155]: Found 27 modes for display 0x04280a40 [27, 0]
	6/22/15 2:47:44.156 PM WindowServer[155]: Received display connect changed for display 0x3f003d
	6/22/15 2:47:44.156 PM WindowServer[155]: Found 1 modes for display 0x003f003d [1, 0]
	6/22/15 2:47:44.157 PM WindowServer[155]: Received display connect changed for display 0x2b282dce
	6/22/15 2:47:44.157 PM WindowServer[155]: Found 1 modes for display 0x2b282dce [1, 0]
	6/22/15 2:47:44.157 PM WindowServer[155]: Display 0x2b282dce changed state to offline
	6/22/15 2:47:44.177 PM WindowServer[155]: Display 0x04280a40: GL mask 0x1; bounds (1920, 1200)[1280 x 800], 27 modes available
	Main, Active, on-line, enabled, built-in, boot, OpenGL-accel, Vendor 610, Model a029, S/N 0, Unit 0, Rotation 0
	UUID 0xe5e926ffff838251e163d61f52faa3e5, ColorSpace { -1569668179 }

Soares / coin_bias_test.py

Last active January 17, 2022 14:32

Let's say you have a fair coin, and can flip it at most 300 times. What percentage of the time will tossing the coin generate a sequence that, at some point along the way, supports the hypothesis "the coin is biased 75% towards heads" 20x as much as it supports the hypothesis "the coin is fair"?

	# Let's say you have a fair coin.
	# Let's say you don't know whether or not it's actually biased towards heads by a certain degree.
	# Let's say you can flip it at most a certain number of times.
	# What's the probability that, at some point along the way, it looks pretty likely to be biased?
	from random import random
	from scipy.misc import comb

	# The actual bias of the coin
	BIAS = 0.5
	# The number of times to flip the coin

Soares / phacker.py

Created July 3, 2016 19:49

Let's say you have a coin that you can flip at most 300 times, and you want to generate a "statistically significant" result, and you're willing to stop as soon as the data looks like it's going your way. How often will you be able to get p < 0.05? p < 0.01?

	# Question: Let's say you have a coin, and you are willing to flip it a bunch of times,
	# and you're going to stop as soon as you get a "statistically significant" bias in favor of heads.
	# Let's say you are willing to flip it at most 300 times.
	# What's the probability that you can get your statistically significant result?
	from sys import stdout
	from random import random
	from scipy.misc import comb

	# The actual bias of the coin
	BIAS = 0.5

Nate Soares

Keybase proof