soodoku

# Load libs 
library(ggplot2)


# Simulate correlated data
R = matrix(cbind(1,.80, .80,1), nrow=2)
U = t(chol(R))

soodoku

The regex patterns in this gist are intended to match any URLs,
including "mailto:foo@example.com", "x-whatever://foo", etc. For a
pattern that attempts only to match web URLs (http, https), see:
https://gist.github.com/gruber/8891611


# Single-line version of pattern:

(?i)\b((?:[a-z][\w-]+:(?:/{1,3}|[a-z0-9%])|www\d{0,3}[.]|[a-z0-9.\-]+[.][a-z]{2,4}/)(?:[^\s()<>]+|\(([^\s()<>]+|(\([^\s()<>]+\)))*\))+(?:\(([^\s()<>]+|(\([^\s()<>]+\)))*\)|[^\s`!()\[\]{};:'".,<>?«»“”‘’]))

soodoku

'''
Text from Searchable pdfs
Scrape Text off Wisconsin Ads pdfs

Uses pyPdf to get text from searchable pdfs. The script is for tailored for getting data 
from Wisconsin Political Ads Database: http://wiscadproject.wisc.edu/Storyboards.

@author: Gaurav Sood

Created on November 02, 2011

soodoku

# Read the data
sf <- read.csv("sf_tenants.csv")

# Recode 
sf$market_rates <- gsub("%", "", sapply(strsplit(sf$market_rate, " / "), "[", 2)) # market_rate
sf$rent_control_rates <- gsub("%", "", sapply(strsplit(sf$rent_control, " / "), "[", 2)) # market_rate

# Ratio of rent_control vs. rent_control 
sf$ratio <- as.numeric(sf$rent_control_rates)/as.numeric(sf$market_rates)

soodoku

"
Basic Text Classifier 		 
	- Takes a csv with a text column, and column of labels
	- Splits into train and test
	- Preprocesses text using tm/bag-of-words, 1/2-order Markov
	- Uses SVM and Lasso

@author: Gaurav Sood								 
 	
"

soodoku

devtools::install_github("abresler/gdeltr2")
devtools::install_github("hafen/trelliscopejs")

library(gdeltr2)
library(dplyr)

asb_ocr <- "Brooklyn Nets"
gkg_codes <-
  get_codes_gkg_themes()
imageweb_codes <- get_gdelt_codebook_ft_api(code_book = "imageweb")

soodoku

Last,First,N,Mode,Date Start,Date End,TotApp,TotDis,TotDK,DemApp,DemDis,DemDK,RepApp,RepDis,RepDK,IndApp,IndDis,IndDK,LibApp,LibDis,LibDK,ModApp,ModDis,ModDK,ConApp,ConDis,ConDK,Source,Link,,,,,,,,,
Bachmann,Michelle,928,Phone,07/15/11,07/17/11,29,45,26,16,58,26,45,27,28,28,47,24,15,64,21,17,53,30,50,24,26,PPP,http://www.publicpolicypolling.com/pdf/2011/PPP_Release_National_720925.pdf,,,,,,,,,
Bachmann,Michelle,700,Phone,10/07/11,10/10/11,27,56,17,15,68,17,43,36,21,23,64,13,11,77,12,14,62,23,49,34,17,PPP,http://www.publicpolicypolling.com/pdf/2011/PPP_Release_US_1011513.pdf,,,,,,,,,
Bachmann,Michelle,700,Phone,12/16/11,12/18/11,30,54,16,22,65,13,41,39,19,26,59,16,21,71,8,20,67,12,42,34,23,PPP,http://www.publicpolicypolling.com/pdf/2011/PPP_Release_National_1220925.pdf,,,,,,,,,
Bloomberg,Michael,707,Phone,11/19/10,11/21/10,19,38,44,24,30,46,12,48,40,19,37,44,28,29,43,24,28,48,9,53,39,PPP,http://publicpolicypolling.blogspot.com/2010/11/americans-not-impressed-with-bloomberg.html,,,,,,,,,
Bloomberg,Michael,700,P

soodoku

Fun Math

Fun Fact 1

• 1 - 1 + 1 - 1 + 1 - 1 + .... = 1/2
• Proof (1) by Luigi Grandi:

S = 1 - 1 + 1 - 1 + 1 - 1 + ...
1 - S = 1 - (1 - 1 + 1 - 1 + 1 - 1 + ...)
= 1 - 1 + 1 - 1 + 1 - 1 + ...

soodoku

# Output = http://gbytes.gsood.com/2013/11/02/the-fairest-of-them-all/

# Uses cces_recode.R here: https://github.com/soodoku/in-n-out/scripts/cces_recode.R

# Plotting the fairest of all media
library(lattice)
png("fairmedia.png")
dotplot(t(t(table(droplevels(cces06$fairmedia[cces06$fairmedia != "Don't know"]), cces06$pid3[cces06$fairmedia != "Don't know"]))/colSums(table(droplevels(cces06$v2112[cces06$fairmedia != "Don't know"]), cces06$pid3[cces06$fairmedia != "Don't know"]))), 
main = "Which network do you think provides the \n fairest coverage of national news?",

soodoku