Continued search, began work on including crime

danielkberry · Nov 20, 2016 · a811fd1 · a811fd1
1 parent a3690ee
commit a811fd1
Show file tree

Hide file tree

Showing 3 changed files with 501 additions and 26 deletions.
diff --git a/ETL.r b/ETL.r
@@ -94,6 +94,18 @@ CTA_counts <- read.csv('CTA_counts.csv')
 
 blocks_raw$CTA_counts <- CTA_counts$x
 
+library(data.table)
+crimes <- fread('../rows.csv')
+
+    t2 <- as.matrix(crime_2009[,c('Latitude', 'Longitude'), with = FALSE])
+tmp <- ddply(blocks_raw, .(GEOID10), function(df) {
+    t1 <- as.matrix(df[1,c('Latitude', 'Longitude')])
+
+    dists <- spDists(t1,t2)
+
+    count <- nrow(crime_2009[])
+})
+
 groceries <- read.csv('food-deserts-master/data/Grocery_Stores_-_2011.csv', stringsAsFactors = FALSE)
 ## drop liquor stores
 groceries <- groceries[grep('liquor', tolower(groceries$STORE.NAME), invert = TRUE),]
@@ -131,7 +143,7 @@ nrow(block_data <- merge(blocks_raw, population, by.x = 'TRACT_BLOC', by.y = 'CE
 
 library(data.table)
 
-crime <- fread('rows.csv')
+crime <- fread('../rows.csv')
 
 library(ggplot2)
 

diff --git a/model.r b/model.r
@@ -1,3 +1,4 @@
+
 ########################
 ## Modeling code file ##
 ########################
@@ -130,6 +131,26 @@ fit_model <- function(i) {
     return(c(i, AIC(model)))
 }
 
+model <- glmer(desert ~ CTA_counts + vacant_counts + Gonorrhea.in.Females + Cancer..All.Sites. + TOTAL.POPULATION + NHAS +
+                   Dependency +Childhood.Lead.Poisoning + Prenatal.Care.Beginning.in.First.Trimester + Gonorrhea.in.Males
+               + NHAM_p + Multiple.Race.. + Stroke..Cerebrovascular.Disease. + Firearm.related + Tuberculosis + NHW_p +
+               Teen.Birth.Rate + No.High.School.Diploma + Lung.Cancer +
+               (1|Neighborhood),
+               data = model_data_scale,
+               family = 'binomial')
+
+summary(model)
+
+in_vars <- c("Gonorrhea.in.Females" , 'Cancer..All.Sites.' , 'TOTAL.POPULATION' , 'NHAS' ,
+                   'Dependency' ,'Childhood.Lead.Poisoning' , 'Prenatal.Care.Beginning.in.First.Trimester' , 'Gonorrhea.in.Males'
+               , 'NHAM_p' , 'Multiple.Race..' , 'Stroke..Cerebrovascular.Disease.' , 'Firearm.related' , 'Tuberculosis' , 'NHW_p' ,
+             'Teen.Birth.Rate' , 'No.High.School.Diploma' , 'Lung.Cancer', 'Colorectal.Cancer', 'HISP_p', 'Below.Poverty.Level', 'Breast.cancer.in.females', 'NHAM')
+out_vars <- setdiff(search_covariates, in_vars)
+
+
+old_aic <- AIC(model)
+
+
 while(TRUE) {
     search_results <- foreach(i=1:length(out_vars), .combine = 'rbind') %dopar% fit_model(i)
     min_aic <- which.min(search_results[,2])
@@ -145,33 +166,35 @@ while(TRUE) {
 }
 print(paste('Final model:', in_vars, collapse = ', '))
 
-model <- glmer(desert ~ CTA_counts + vacant_counts + Gonorrhea.in.Females + Cancer..All.Sites. + TOTAL.POPULATION + NHAS +
-                   Dependency +Childhood.Lead.Poisoning + Prenatal.Care.Beginning.in.First.Trimester + Gonorrhea.in.Males
-               + NHAM_p + Multiple.Race.. + Stroke..Cerebrovascular.Disease. + Firearm.related + Tuberculosis + NHW_p +
-               Teen.Birth.Rate + No.High.School.Diploma + Lung.Cancer +
-               (1|Neighborhood),
-               data = model_data_scale,
-               family = 'binomial')
 
-summary(model)
 
-in_vars <- c("Gonorrhea.in.Females" , 'Cancer..All.Sites.' , 'TOTAL.POPULATION' , 'NHAS' ,
-                   'Dependency' ,'Childhood.Lead.Poisoning' , 'Prenatal.Care.Beginning.in.First.Trimester' , 'Gonorrhea.in.Males'
-               , 'NHAM_p' , 'Multiple.Race..' , 'Stroke..Cerebrovascular.Disease.' , 'Firearm.related' , 'Tuberculosis' , 'NHW_p' ,
-             'Teen.Birth.Rate' , 'No.High.School.Diploma' , 'Lung.Cancer')
-out_vars <- setdiff(search_covariates, in_vars)
+## model <- glmer(desert ~ CTA_counts + vacant_counts + Gonorrhea.in.Females + Cancer..All.Sites. + TOTAL.POPULATION + NHAS +
+##                    Dependency +Childhood.Lead.Poisoning + Prenatal.Care.Beginning.in.First.Trimester + Gonorrhea.in.Males
+##                + NHAM_p + Multiple.Race.. + Stroke..Cerebrovascular.Disease. + Firearm.related + Tuberculosis + NHW_p +
+##                Teen.Birth.Rate + No.High.School.Diploma + Lung.Cancer +
+##                (1|Neighborhood),
+##                data = model_data_scale,
+##                family = 'binomial')
 
+## summary(model)
 
-old_aic <- AIC(model)
+## in_vars <- c("Gonorrhea.in.Females" , 'Cancer..All.Sites.' , 'TOTAL.POPULATION' , 'NHAS' ,
+##                    'Dependency' ,'Childhood.Lead.Poisoning' , 'Prenatal.Care.Beginning.in.First.Trimester' , 'Gonorrhea.in.Males'
+##                , 'NHAM_p' , 'Multiple.Race..' , 'Stroke..Cerebrovascular.Disease.' , 'Firearm.related' , 'Tuberculosis' , 'NHW_p' ,
+##              'Teen.Birth.Rate' , 'No.High.School.Diploma' , 'Lung.Cancer')
+## out_vars <- setdiff(search_covariates, in_vars)
 
-search_results <- foreach(i=1:length(out_vars), .combine = 'rbind') %dopar% fit_model(i)
-min_aic <- which.min(search_results[,2])
-if (min_aic < old_aic) {
-    print(paste('ADDING:',out_vars[min_aic]))
-    in_vars <- c(in_vars, out_vars[min_aic])
-    out_vars <- setdiff(out_vars, out_vars[min_aic])
-    print(search_results[min_aic,])
-}
-else {
-    break
-}
+
+## old_aic <- AIC(model)
+
+## search_results <- foreach(i=1:length(out_vars), .combine = 'rbind') %dopar% fit_model(i)
+## min_aic <- which.min(search_results[,2])
+## if (min_aic < old_aic) {
+##     print(paste('ADDING:',out_vars[min_aic]))
+##     in_vars <- c(in_vars, out_vars[min_aic])
+##     out_vars <- setdiff(out_vars, out_vars[min_aic])
+##     print(search_results[min_aic,])
+## }
+## else {
+##     break
+## }