MichaelChirico

install.packages('RPushbullet')
# get key here:
#  https://www.pushbullet.com/#settings/account
# note: this will prompt for a default device
RPushbullet::pbSetup('keep_this_private')

library(RPushbullet)
#that's it!
pbPost('note', "IT'S", 'ALIVE')

MichaelChirico

MichaelChirico

feb28 = feb28 = structure(list(x_coordina = c(45.590501, 45.567097, 45.518155, 
45.509379, 45.521205, 45.568948, 45.525444, 45.527497, 45.537308, 
45.522968, 45.537873, 45.506088, 45.534096, 45.481038, 45.521164, 
45.510095, 45.518234, 45.518234, 45.5263, 45.511535, 45.535158, 
45.515665, 45.516012, 45.479503, 45.515318, 45.520765, 45.515158, 
45.446186, 45.514426, 45.466002, 45.553682, 45.532582, 45.51153, 
45.497917, 45.515003, 45.503943, 45.530414, 45.510336, 45.533919, 
45.522286, 45.534656, 45.5322, 45.481887, 45.574603, 45.509471, 
45.543617, 45.584307, 45.537896, 45.518114, 45.532731, 45.512403, 
45.497119, 45.441898, 45.533295, 45.469783, 45.563008, 45.515077, 

MichaelChirico

library(data.table)
whatsapp_raw = readLines('~/Downloads/WhatsApp Chat with PhDelphia.txt')

#have to deal with multi-line messages :\
idx = grepl('^1?[0-9]/', whatsapp_raw)

idxrle = rle(idx)
bdpts = cumsum(idxrle$lengths)
for (ii in seq_len(length(idxrle$values))) {
  if (!idxrle$values[ii]) {

MichaelChirico

library(gh)
library(data.table)

user = 'Rdatatable'
repo = 'data.table'

# GET URL for 100 open issues at a time
issue_query_fmt = "/repos/%s/%s/issues?state=open&per_page=100&page=%d"
issues = list()
page = 1L

MichaelChirico

train_set,delx,dely,alpha,eta,lt,theta,k,l1,l2,kde.bw,kde.lags,kde.win,pei,pai
train_13,181.510032358347,620.980401238194,0,0.956507977447473,7,2.06174959123186,100,0,0,615.388604282634,1,11,0.0952380952380952,24.4244783950786
train_13,181.510032358347,620.980401238194,0,0.956507977447473,7,2.06174959123186,100,0,1e-05,615.388604282634,1,11,0.0238095238095238,6.10611959876964
train_13,181.510032358347,620.980401238194,0,0.956507977447473,7,2.06174959123186,100,0,5e-05,615.388604282634,1,11,0,0
train_13,181.510032358347,620.980401238194,0,0.956507977447473,7,2.06174959123186,100,0,1e-04,615.388604282634,1,11,0.0238095238095238,6.10611959876964
train_13,181.510032358347,620.980401238194,0,0.956507977447473,7,2.06174959123186,100,0,5e-04,615.388604282634,1,11,0.0238095238095238,6.10611959876964
train_13,181.510032358347,620.980401238194,0,0.956507977447473,7,2.06174959123186,100,0,0.001,615.388604282634,1,11,0.0119047619047619,3.05305979938482
train_13,181.510032358347,620.980401238194,0,0.956507977447473,7,2.06

MichaelChirico

setwd('~/Downloads')
library(rgdal)
library(rgeos)
library(data.table)

#ground truth crimes file
obs_crimes = readOGR('PPB Data', 'NIJ2017_MAR01_May31')
obs_crimes$occ_date = as.IDate(obs_crimes$occ_date, format = '%Y/%m/%d')

#get indices corresponding to each crime category

MichaelChirico

\begin{itemized}

\item $L(x) = f(x) g(h(q(r(x))))$, where $f(x) = \frac{1}{\sigma \sqrt{2 \pi}} \frac{1}{x}$, 
  $g(x) = e^x$, $h(x) = -\frac12 x^2$, $q(x) = \frac{x - \mu}{\sigma}$, and $r(x) = \log(x)$.
  
\item \texttt{f = function(x) exp(-.5*log(x)^2)/x/sqrt(2*pi)}

\item \texttt{curve(f, 0, 5)}

\item No. We can see this on the graph. Analytically, both $L(.136)$ and $L(.999)$ are roughly equal to .4.

MichaelChirico

library(rvest)
library(data.table)

# downloaded individually... website throttles scraping >:(
#   URL stub for year YYYY is 
#   http://www.hockeyfights.com/leaders/teams/1/regYYYY
pgs = list.files('~/Desktop/nhl_fights',
                 full.names = TRUE, pattern = 'html')
names(pgs) = gsub('.*([0-9]{4}).*', '\\1', pgs)

MichaelChirico

# load tools for manipulating data
library(data.table)
# read data (press TAB to enable path completion)
prop = fread('~/Downloads/opa_properties_public.csv')

# first,  subset to 19125
## exploring: is zip code consistently formatted? (no)
prop[ , table(nchar(zip_code))]
## mainly looks like there are 6 formats:
##   1. 1912 [typo; 2 cases]