daynebatten

function calculate() {
  for (i=0; i<=99; i++) {
    probs[i]=[];
    
    for (j=0; j<=99; j++) {
      probs[i][j]=[];
      
      for (k=0; k<=99; k++){
        probs[i][j][k]=.5;
      }

daynebatten

AppTitle "Mouse Distance"

file=ReadFile("distance.ini")
Global custom=ReadLine(file)
Global width=ReadLine(file)
Global height=ReadLine(file)
Global writedelay=ReadLine(file)
Global inches#=ReadLine(file)
Global feet=ReadLine(file)
Global miles=ReadLine(file)

daynebatten

female	time	churned
0	273	0
0	360	0
0	8	0
0	211	0

daynebatten

# Use the survival library

library(survival)

# Read in the NetLixx data... yes, I use Windows.

net_lixx <- read.csv('C:/Users/dayne/Desktop/NetLixx.csv')

# Create a "survival object" for each observation, using time and churn data.

daynebatten

# Run basic setup again

library(survival)
net_lixx <- read.csv('C:/Users/dayne/Desktop/NetLixx.csv')
net_lixx$survival <- Surv(net_lixx$time, net_lixx$churned == 1)

# Fit survival curves based on the female binary

fit <- survfit(survival ~ female, data = net_lixx)

daynebatten

Call:
survdiff(formula = survival ~ female, data = net_lixx)

            N Observed Expected (O-E)^2/E (O-E)^2/V
female=0 6722      830      778      3.48      12.4
female=1 2532      251      303      8.94      12.4

 Chisq= 12.4  on 1 degrees of freedom, p= 0.000422 

daynebatten

female    age    coupon    churned    time
0         64     0         1          158
0         22     0         0          25
0         37     0         0          150
0         55     0         0          311

daynebatten

# Load the survival library

library(survival)

# Read in the data and create a "survival object"

netlixx_cox <- read.csv('C:/Users/dayne/Desktop/NetLixxCox.csv')
netlixx_cox$survival <- Surv(netlixx_cox$followtime, netlixx_cox$churn == 1)

# Run a cox regression of survival on female, age, and coupon status

daynebatten

coxph(formula = survival ~ female + age + coupon, data = netlixx_cox)


           coef exp(coef) se(coef)      z       p
female -0.03749     0.963  0.09098 -0.412 6.8e-01
age    -0.00561     0.994  0.00176 -3.191 1.4e-03
coupon  0.58709     1.799  0.08104  7.245 4.3e-13

Likelihood ratio test=58.3  on 3 df, p=1.37e-12  n= 5124, number of events= 743 

daynebatten

# Run a test for non-proportional hazards on our cox regression results

cox.zph(results)