-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathMasterCodeofCashmereProject_plm_robustness_bedrooms_interactions (1).Rmd
218 lines (167 loc) · 7.71 KB
/
MasterCodeofCashmereProject_plm_robustness_bedrooms_interactions (1).Rmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
---
title: "Robustness Check with subsample of houses with two or more bedrooms II"
author: "Peng Sun"
date: "10/03/2022"
output:
pdf_document: default
html_document: default
---
```{r load packages, message=FALSE, warning=FALSE}
#Load packages
#install.packages("clubSandwich")
library(readr)
library(did)
library(fixest)
library(DRDID)
library(tidyverse)
library(plm)
library(stargazer)
library(sandwich)
library(lmtest)
library(jtools)
library(clubSandwich)
```
```{r,message=FALSE, warning=FALSE}
#read data
setwd("~/R")
df2012<-read_csv("~/R/df2012_20210920.csv")
```
```{r}
# set levels of timedummy variables so that Y2012 will be used as
names(df2012)
unique(df2012$timedummy)
```
```{r, message=FALSE, warning=FALSE}
df2012$timedummy<-fct_relevel(df2012$timedummy,levels="Y2012","Y2013","Y2014","Y2015","Y2016","Y2017","janapr18","t1","t2")
#set levels of groups dummies so that Group C will be used as base
df2012$groups<-fct_relevel(df2012$groups, levels="C","A","B","D")
# set levels of period dummies so that t0 period will be used as base
df2012$periods<-fct_relevel(df2012$periods, levels="t0","t1","t2")
df2012<-df2012%>%select(-1)
```
```{r}
df2012<-df2012 %>%
mutate(timedummy2=case_when(year==2012~"Y2012",
year==2013~"Y2013",
year==2014~"Y2014",
year==2015~"Y2015",
year==2016~"Y2016",
year==2017~"Y2017",
janapr18==1~"Y2018janapr",
t1==1~"Yt1",
t2==1~"Yt2"))
unique(df2012$timedummy2)
```
# Robustness check with Control of interaction terms among bedroom#*period, bedroom#*group, bedroom*period*group
By adding interaction terms:
"A*bedrooms_min+B*bedrooms_min+D*bedrooms_min+t1*bedrooms_min+t2*bedrooms_min+A*t1*bedrooms_min+B*t1*bedrooms_min+D*t1*bedrooms_min+A*t2*bedrooms_min+B*t2*bedrooms_min+D*t2*bedrooms_min"
## Create the subset of houses with two or more bedrooms
```{r}
unique(df2012$bedrooms_min)
```
```{r}
# check how many obs with one bedroom, zero bedroom or NA
df2012_2beds<- df2012 %>% filter(bedrooms_min<=1|bedrooms_min =="NA")
dim(df2012_2beds) # 168 obs to be dropped for the robustness check
```
```{r}
df2012_2morebeds<- df2012 %>% filter(bedrooms_min>1)
dim(df2012_2morebeds)
```
```{r}
df2012<-df2012_2morebeds
```
## Create the imputated dataset (used in Col 2 and 4)
```{r}
# use the mean land_area and floor_area for imputation
mean(df2012$land_area, na.rm="true")
```
```{r}
mean(df2012$floor_area)
# only keep three significant figures
```
```{r}
df2012_mutated <- df2012
df2012_mutated$contour=replace_na(df2012_mutated$contour,"Unreported")
df2012_mutated$deck=replace_na(df2012_mutated$deck,"Unreported")
df2012_mutated$roof_material=replace_na(df2012_mutated$roof_material,"Unreported")
df2012_mutated$wall_material=replace_na(df2012_mutated$wall_material,"Unreported")
df2012_mutated$building_age=replace_na(df2012_mutated$building_age,"Unreported")
df2012_mutated$land_area=replace_na(df2012_mutated$land_area,0.069)
df2012_mutated$floor_area=replace_na(df2012_mutated$floor_area,136)
```
```{r}
garage=df2012_mutated$garages_mainroof
```
```{r}
# replace msising values with Unreported
garage[is.na(garage)]<-"Unreported"
```
```{r}
which(is.na(garage))
```
```{r}
df2012_mutated$garages_mainroof<-garage
```
```{r}
bedrooms_min=df2012_mutated$bedrooms_min
# replace msising values with Unreported
bedrooms_min[is.na(bedrooms_min)]<-"Unreported"
#head(bedrooms_min)
df2012_mutated$bedrooms_min<- bedrooms_min
```
## column 1
```{r}
col1_bedrm<-plm(lp~t1+t2+A+B+D+t1*A+t2*A+t1*B+t2*B+t1*D+t2*D+A*bedrooms_min+B*bedrooms_min+D*bedrooms_min+t1*bedrooms_min+t2*bedrooms_min+A*t1*bedrooms_min+B*t1*bedrooms_min+D*t1*bedrooms_min+A*t2*bedrooms_min+B*t2*bedrooms_min+D*t2*bedrooms_min,data=df2012, model="pooling", index=c("houseid","year") )
summary(col1_bedrm)
```
```{r}
coefplot(col1_bedrm,keep=c("t1:A","t2:A","t1:B","t2:B","t1:D","t2:D"), ylim=c(-0.4,0.4), axis.text.x = element_text(size=14))
```
## Column 2
```{r}
col2_bedrm<-lm(lp~t1+t2+t1*A+t2*A+t1*B+t2*B+t1*D+t2*D+floor_area+land_area+ bedrooms_min +as.factor(building_age)+ as.factor(deck)+ as.factor(wall_material)
+as.factor(roof_material)+as.factor(garages_mainroof)+ as.factor(contour)+as.factor(timedummy2) +as.factor(suburb)+A*bedrooms_min+B*bedrooms_min+D*bedrooms_min+t1*bedrooms_min+t2*bedrooms_min+A*t1*bedrooms_min+B*t1*bedrooms_min+D*t1*bedrooms_min+A*t2*bedrooms_min+B*t2*bedrooms_min+D*t2*bedrooms_min,data=df2012_mutated)
summary(col2_bedrm)
```
```{r}
coefplot(col2_bedrm,keep=c("t1:A","t2:A","t1:B","t2:B","t1:D","t2:D"), ylim=c(-0.4,0.4), axis.text.x = element_text(size=14))
```
## Column 3
```{r}
col3_bedrm<-plm(lp~yr2012+yr2013+yr2014+yr2015+yr2016+yr2017+t1+t2+A+B+D+t1*A+t2*A+t1*B+t2*B+t1*D+t2*D+A*bedrooms_min+B*bedrooms_min+D*bedrooms_min+yr2012*bedrooms_min+yr2013*bedrooms_min+yr2014*bedrooms_min+yr2015*bedrooms_min+yr2016*bedrooms_min+yr2017*bedrooms_min+t1*bedrooms_min+t2*bedrooms_min+A*t1*bedrooms_min+B*t1*bedrooms_min+D*t1*bedrooms_min+A*t2*bedrooms_min+B*t2*bedrooms_min+D*t2*bedrooms_min,data=df2012, model="within",effect="individual", index=c("houseid","year"))
summary(col3_bedrm)
```
```{r}
coefplot(col3_bedrm,keep=c("t1:A","t2:A","t1:B","t2:B","t1:D","t2:D"), ylim=c(-0.4,0.4), axis.text.x = element_text(size=14))
```
## Column 4
```{r}
col4_bedrm<-plm(lp~yr2012+yr2013+yr2014+yr2015+yr2016+yr2017+yr2012*A+yr2013*A+yr2014*A+yr2015*A+yr2016*A+yr2017*A+yr2012*B+yr2013*B+yr2014*B+yr2015*B+yr2016*B+yr2017*B+yr2012*D+yr2013*D+yr2014*D+yr2015*D+yr2016*D+yr2017*D+t1+t2+A+B+D+t1*A+t2*A+t1*B+t2*B+t1*D+t2*D+floor_area+land_area+bedrooms_min+as.factor(building_age)+ as.factor(deck)+ as.factor(wall_material)
+as.factor(roof_material)+as.factor(garages_mainroof)+ as.factor(contour)+as.factor(suburb)+yr2014*bedrooms_min+yr2015*bedrooms_min+yr2016*bedrooms_min+yr2017*bedrooms_min+t1*bedrooms_min+t2*bedrooms_min+A*t1*bedrooms_min+B*t1*bedrooms_min+D*t1*bedrooms_min+A*t2*bedrooms_min+B*t2*bedrooms_min+D*t2*bedrooms_min,data=df2012_mutated,model="pooling", index=c("houseid","year"))
summary(col4_bedrm)
```
```{r}
coefplot(col4_bedrm,ylim=c(-0.4,0.4), keep=c("yr2012:A","yr2013:A","yr2014:A","yr2015:A","yr2016:A","yr2017:A","A:t1", "A:t2"), axis.text.x = element_text(size=14))
```
```{r}
coefplot(col4_bedrm,keep=c("yr2012:B","yr2013:B","yr2014:B","yr2015:B","yr2016:B","yr2017:B", "B:t1","B:t2"), axis.text.x = element_text(size=14))
```
```{r}
coefplot(col4_bedrm, keep=c("yr2012:D","yr2013:D","yr2014:D","yr2015:D","yr2016:D","yr2017:D", "D:t1", "D:t2"), axis.text.x = element_text(size=14))
```
## Column 5
```{r}
col5_bedrm<-plm(lp~yr2012+yr2013+yr2014+yr2015+yr2016+yr2017+yr2012*A+yr2013*A+yr2014*A+yr2015*A+yr2016*A+yr2017*A+yr2012*B+yr2013*B+yr2014*B+yr2015*B+yr2016*B+yr2017*B+yr2012*D+yr2013*D+yr2014*D+yr2015*D+yr2016*D+yr2017*D+t1+t2+A+B+D+t1*A+t2*A+t1*B+t2*B+t1*D+t2*D+yr2014*bedrooms_min+yr2015*bedrooms_min+yr2016*bedrooms_min+yr2017*bedrooms_min+t1*bedrooms_min+t2*bedrooms_min+A*t1*bedrooms_min+B*t1*bedrooms_min+D*t1*bedrooms_min+A*t2*bedrooms_min+B*t2*bedrooms_min+D*t2*bedrooms_min,data=df2012, model="within",effect="individual", index=c("houseid","year"))
summary(col5_bedrm)
```
#Coef graph of column 5
```{r}
coefplot(col5_bedrm, ylim=c(-0.4,0.4),keep=c("yr2012:A","yr2013:A","yr2014:A","yr2015:A","yr2016:A","yr2017:A","janapr18:A", "A:t1","A:t2"), axis.text.x = element_text(size=14))
```
```{r}
coefplot(col5_bedrm,keep=c("yr2012:B","yr2013:B","yr2014:B","yr2015:B","yr2016:B","yr2017:B", "B:t1", "B:t2"), axis.text.x = element_text(size=14))
```
```{r}
coefplot(col5_bedrm, keep=c("yr2012:D","yr2013:D","yr2014:D","yr2015:D","yr2016:D","yr2017:D","D:t1","D:t2"), axis.text.x = element_text(size=14))
```