Homework 5
Tobias Kuhn
March 5, 2018
library(haven)
library(readr)
library(dplyr)##
## Attaching package: 'dplyr'## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionlibrary(ggplot2)
library(broom)
library(car)##
## Attaching package: 'car'## The following object is masked from 'package:dplyr':
##
## recodelibrary(stargazer)##
## Please cite as:## Hlavac, Marek (2018). stargazer: Well-Formatted Regression and Summary Statistics Tables.## R package version 5.2.1. https://CRAN.R-project.org/package=stargazerlibrary(survey)## Loading required package: grid## Loading required package: Matrix## Loading required package: survival##
## Attaching package: 'survey'## The following object is masked from 'package:graphics':
##
## dotchartlibrary(questionr)
library(pander)#importing RECS 2015 data
recs2015v2<-read.csv("C:/Users/tobik/Downloads/recs2015_public_v2.csv")Creating a subset without missing values:
myrecs3<-subset(recs2015v2,ZMICRO=1) #has a microwave
myrecs3<-subset(myrecs3,ZOVEN=1) #has an oven
myrecs3<-subset(myrecs3,PROTHERM>-1) # has programmable thermostat question answered / appliccable
myrecs3<-subset(myrecs3,ZSTOVE=1) #has a stove
myrecs3<-subset(myrecs3,EQUIPAGE>0) #has an heater
myrecs3<-subset(myrecs3,AGECENAC>0) #has an AC question answered / appliccable
myrecs3<-subset(myrecs3,ZLGTINLED=1) #answered LED light bulb questionRecoding missing values as zeros for calcuation of cooking methods (-2 indicates “NA”, thus it is effectively a not used case, or 0)
myrecs3$COOKTUSE<-recode(myrecs3$COOKTUSE, recodes = "-2:0=0")
myrecs3$OVENUSE<-recode(myrecs3$OVENUSE, recodes = "-2:0=0")
myrecs3$SEPCOOKTUSE<-recode(myrecs3$SEPCOOKTUSE, recodes = "-2:0=0")
myrecs3$SEPOVENUSE<-recode(myrecs3$SEPOVENUSE, recodes = "-2:0=0")
myrecs3$AMTMICRO<-recode(myrecs3$AMTMICRO, recodes = "-2:0=0")Calculating the value to be used for binary indicator
Microwave use over Stove/Cook-top use:
myrecs3$mycooks<-myrecs3$COOKTUSE
myrecs3$mycooks<-myrecs3$mycooks+myrecs3$OVENUSE
myrecs3$mycooks<-myrecs3$mycooks+myrecs3$SEPCOOKTUSE
myrecs3$mycooks<-myrecs3$mycooks+myrecs3$SEPOVENUSE
myrecs3$mycooks<-myrecs3$mycooks-myrecs3$AMTMICRO
myrecs3$mycooks<-recode(myrecs3$mycooks, recodes="0:396=1;-99:-1=0")
myrecs3$mycooks<-recode(myrecs3$mycooks, recodes="1=2;-99:0=3") #flipping 1
myrecs3$mycooks<-recode(myrecs3$mycooks, recodes="2=0;-99:3=1") #flipping 2
## 0 = prefers cooking with stove / stovetop
## 1 = uses microwave at least as oftenAge of heating equipment:
myrecs3$heating<-myrecs3$EQUIPAGE
myrecs3$heating<-recode(myrecs3$heating, recodes="1:3=1;5:42=0")
## 0 = 10 years and older equipment
## 1 = less than 9 years old = 1Programmable Thermostat for heating: Thermost is coded as PROTHERM = 1 for having one and = 0 for not having one
Age of AC equipment:
myrecs3$cooling<-myrecs3$AGECENAC
myrecs3$cooling<-recode(myrecs3$cooling, recodes="1:3=1;5:42=0")
## 0 = 10 years and older equipment
## 1 = less than 9 years old = 1Uses at least most LED light fixtures:
myrecs3$LEDbulbs<-myrecs3$LGTINLED
myrecs3$LEDbulbs<-recode(myrecs3$LEDbulbs, recodes="1:2=1;3:4=0")
## 0 = about half to none
## 1 = Most or All LED bublsPrinciple Components Eco-Friendly Energy User
Components Selected:
- Uses at least most LED light fixtures
- Has a Programmable Heater Thermostat
- Age of Heating Equipment (<10y / 10y+)
- Age of AC Equipment (<10y / 10y+)
- Prefers Microwave Use
The variables are: mycooks heating PROTHERM cooling LEDbulbs
The scores do not need to be standardized because they are all binary.
myrecs3.pc<-prcomp(~LEDbulbs+PROTHERM+heating+cooling+mycooks,data=myrecs3, center=T,retx=T)
summary(myrecs3.pc)## Importance of components:
## PC1 PC2 PC3 PC4 PC5
## Standard deviation 0.6513 0.4995 0.4691 0.2819 0.21944
## Proportion of Variance 0.4153 0.2443 0.2154 0.0778 0.04715
## Cumulative Proportion 0.4153 0.6596 0.8751 0.9528 1.00000myrecs3.pc$rotation## PC1 PC2 PC3 PC4 PC5
## LEDbulbs 0.0007640987 -0.004936221 0.01662451 0.0428534040 0.998930559
## PROTHERM 0.1665177163 0.072674156 0.98319585 -0.0082016587 -0.015779058
## heating 0.6989498882 -0.017298808 -0.12246090 -0.7036861672 0.031605543
## cooling 0.6954152884 -0.016846168 -0.11108560 0.7091697657 -0.029189333
## mycooks -0.0117429476 -0.997051182 0.07558354 -0.0005831635 -0.006150819hist(myrecs3.pc$x[,1])
hist(myrecs3.pc$x[,2])
Correlations between the first 2 components:
cor(myrecs3.pc$x[,1:2])## PC1 PC2
## PC1 1.00000e+00 -2.47691e-15
## PC2 -2.47691e-15 1.00000e+00scores<-data.frame(myrecs3.pc$x)
scores$name<-rownames(myrecs3.pc$x)
myrecs3$name<-rownames(myrecs3)
myrecs3<-merge(myrecs3, scores, by.x="name", by.y="name", all.x=F)
tail(names(myrecs3), 20)## [1] "BRRWT86" "BRRWT87" "BRRWT88" "BRRWT89" "BRRWT90" "BRRWT91"
## [7] "BRRWT92" "BRRWT93" "BRRWT94" "BRRWT95" "BRRWT96" "mycooks"
## [13] "heating" "cooling" "LEDbulbs" "PC1" "PC2" "PC3"
## [19] "PC4" "PC5"Correlations: (LEDbulbs PROTHERM mycooks heating cooling )
round(cor(myrecs3[,c("LEDbulbs","PROTHERM","heating","cooling","mycooks")], method = "spearman"), 3)## LEDbulbs PROTHERM heating cooling mycooks
## LEDbulbs 1.000 0.027 -0.010 0.008 0.011
## PROTHERM 0.027 1.000 0.097 0.103 -0.011
## heating -0.010 0.097 1.000 0.681 -0.005
## cooling 0.008 0.103 0.681 1.000 -0.005
## mycooks 0.011 -0.011 -0.005 -0.005 1.000Discussion
This exercise can only be seen as an programming task because that is most of what is left. The only strong correlation is newer Central Heat and newer Central AC, which understandably correlated since they are - when new - most of the time the same unit. The concept was that people who are eco-friendly would have LED light bulbs, a programmable thermostat (which I found out is always attached to the hearter - with the exception of one case where it is only for the AC), have a newer heater or AC (less than 10 years old) and prefer the microwave over oven/stove use to save energy by cooking once and heating up withv more energy efficient microwaves.