class_scripts/class_preps.R
In reuning/POL306: POL 306 Applied Research Methods at Miami
setwd("~/Dropbox/Teaching_Kevin/POL306/class_scripts/")
df <- read.csv("qog_oecd_cs_jan21.csv")
df$gea_ea1524f
df$wdi_fertility
mod <- lm(cpds_vt ~ cpds_enps, data=df)
summary(mod)
png("turnout_v_party.png", height=4, width=6, units="in", res=200)
plot(y=df$cpds_vt, x=df$cpds_enps, xlab="Effective Number of Parties",
ylab="Voter Turnout", main="Turnout vs Number of Parties")
abline(mod)
dev.off()
stargazer::stargazer(mod, se =NA,
covariate.labels = c("Effective Number of Parties", "Intercept"),
dep.var.labels = "Turnout",
no.space = T, omit.stat = c("rsq", "adj.rsq", "f", "ser"), type="html")
df <- read.csv("https://www.kevinreuning.com/assets/state_cig.csv")
mod <- lm(cig_tax~liberal, data=df)
summary(mod)
mod <- lm(cig_tax~female_leg, data=df)
summary(mod)
nobs(mod)
mod <- lm(healthspendpc~female_leg, data=df)
summary(mod)
nobs(mod)
confint(mod, level=0.99)
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/resid.pdf",
height=8, width=8)
par(mfrow=c(2,2))
plot(rnorm(1000), ylab="Residual", main="Good!")
plot(rnorm(1000, sd=seq(.5,3, length.out = 1000)), ylab="Residual", main="Heteroskedasticity")
set.seed(1)
tmp <- rnorm(1000)
tmp2 <- tmp + .25*cumsum(tmp)
mean(tmp2)
plot(tmp2, ylab="Residual",main="Autocorrelation")
dev.off()
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/resid2.pdf",
height=6, width=8)
par(mfrow=c(1,2))
x <- rnorm(1000)
y <- 2*x + rnorm(1000)
mod <- lm(y~x)
plot(y=mod$residuals, x=mod$fitted.values, xlab='Predicts', ylab='Residuals', main='Good')
x <- rnorm(1000)
z <- -x + rnorm(1000, 2)
y <- 2*x + rnorm(1000,sd=exp(x))
mod <- lm(y~x)
plot(y=mod$residuals, x=mod$fitted.values, xlab='Predicts', ylab='Residuals', main='Heteroskedasticity')
dev.off()
mod <- lm(cig_tax~female_leg, data=df)
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/resid3.pdf",
height=6, width=8)
par(mfrow=c(1,2))
plot(mod$residuals, ylab='Residuals')
plot(y=mod$residuals, x=mod$fitted.values, xlab='Predicts', ylab='Residuals')
dev.off()
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/null_dist.pdf",
height=6, width=8)
plot(y=dnorm(seq(-.05,.05,by=.001),
sd=sqrt(diag(vcov(mod)))[2] ),
x=seq(-.05,.05, by=.001), type='l', ylab='Density',
xlab='Coefficient Value',
main='Null Distribution')
abline(v=coef(mod)[2], lty=2)
text(x=coef(mod)[2], y=50, pos=3, labels="Estimated Value",
srt=90)
dev.off()
pdf("C:/Users/reunink/Dropbox/Teaching_Kevin/Spring_2021/POL_306/Lectures/images/crit_vals.pdf",
height=6, width=8)
plot(y=dnorm(seq(-3,3,by=.1)),
x=seq(-3,3, by=.1), type='l',
ylab='Density',
xlab="Value",
main="t-distribution")
plot(y=dnorm(seq(-3,3,by=.1)),
x=seq(-3,3, by=.1), type='l',
ylab='Density',
xlab="Value",
main="t-distribution")
polygon(x=rev(c(-1.644854,seq(-1.644854, -3, by=-.1))),
y=rev(c(0.004992899,dnorm(seq(1.644854, 3, by=.1)))),
col='gray')
abline(v=-1.644854)
arrows(y1=dnorm(2), x=-2,
y0=0.15, x0=-2.25
)
text(x=-2.25, y=.15,
label='0.05 Area', pos=3)
text(x=-1.644854, pos=3, y=.2,
label='-1.644854 Crit Value', xpd=T,
srt=270)
plot(y=dnorm(seq(-3,3,by=.1)),
x=seq(-3,3, by=.1), type='l',
ylab='Density',
xlab="Value",
main="t-distribution")
polygon(x=c(seq(-3, -2.326348, length.out = 20), -2.326348),
y=c(dnorm(seq(-3, -2.326348, length.out = 20)), 0.004992899),
col='gray')
abline(v=-2.326348)
arrows(y1=dnorm(-2.5), x1=-2.5,
y0=0.15, x0=-2.75
)
text(x=-2.75, y=.15,
label='0.01 Area', pos=3)
text(x=-2.326348, pos=3, y=.2,
label='-2.326348 Crit Value', xpd=T,
srt=270)
plot(y=dnorm(seq(-3,3,by=.1)),
x=seq(-3,3, by=.1), type='l',
ylab='Density',
xlab="Value",
main="t-distribution")
polygon(x=c(seq(-3, -1.959964, length.out = 20), -1.959964),
y=c(dnorm(seq(-3, -1.959964, length.out = 20)), 0.004992899),
col='gray')
abline(v=-1.959964)
arrows(y1=dnorm(-2.25), x1=-2.25,
y0=0.15, x0=-2.75
)
text(x=-2.75, y=.15,
label='0.025 Area', pos=3)
text(x=-1.959964, pos=3, y=.2,
label='-1.959964 Crit Value', xpd=T,
srt=270)
polygon(x=c(1.959964,seq(1.959964, 3, by=.1)),
y=c(0.004992899,dnorm(seq(1.959964, 3, by=.1))),
col='gray')
abline(v=1.959964)
arrows(y1=dnorm(2.25), x1=2.25,
y0=0.15, x0=2.75
)
text(x=2.65, y=.15,
label='0.025 Area', pos=3)
text(x=1.959964, pos=3, y=.2,
label='1.959964 Crit Value', xpd=T,
srt=270)
dev.off()
pnorm(1.68)
qnorm(0.025)
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/null_dist2.pdf",
height=6, width=8)
plot(y=dnorm(seq(-.05,.05,by=.001),
sd=sqrt(diag(vcov(mod)))[2] ),
x=seq(-.05,.05, by=.001), type='l', ylab='Density',
xlab='Coefficient Value',
main='Null Distribution')
abline(v=0.005, lty=2)
text(x=0.005, y=50, pos=3, labels="Estimated Value",
srt=90)
dev.off()
df[df$state=="Ohio",]
df[df$state=="Indiana",]
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/corr_biv.pdf",
height=8, width=8)
plot(df$liberal, df$female_leg, xlab="% Liberal Voters", ylab="% Women Leg")
dev.off()
cor(df$liberal, df$female_leg, use="complete.obs")
mod <- lm(cig_tax~female_leg + liberal, data=df)
summary(mod)
nobs(mod)
confint(mod)
mod <- lm(cig_tax~female_leg , data=df)
summary(mod)
nobs(mod)
mod <- lm(cig_tax~liberal , data=df)
summary(mod)
nobs(mod)
mod <- lm(healthspendpc~cig_tax , data=df)
summary(mod)
nobs(mod)
confint(mod)
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/assums_ex.pdf",
height=8, width=12)
par(mfrow=c(1,2))
plot(mod$residuals)
plot(y=mod$residuals, x=mod$fitted.values)
dev.off()
full_df <- read.csv("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/COSP_Limited.csv")
tmp <- full_df[,c("z_cigarette_taxes", "pctfemaleleg",
"liberal", "conservative", "state", "year",
"democrat",
"hou_chamber")]
tmp <- tmp[tmp$year > 2000,]
tmp <- tmp[tmp$state!="Delaware",]
names(tmp)[1:2] <- c("cig_taxes", "female_leg")
mod1 <- lm(cig_taxes~liberal+hou_chamber, data=tmp)
mod2 <- lm(cig_taxes~hou_chamber, data=tmp)
mod3 <- lm(cig_taxes~liberal, data=tmp)
summary(mod1)
stargazer::stargazer(mod1, mod2, mod3)
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/multico.pdf",
height=8, width=8)
plot(tmp$liberal, tmp$hou_chamber, xlab="Percentage Liberals",
ylab="Legislative Ideology")
dev.off()
cor(tmp$liberal, tmp$hou_chamber, use='complete.obs')
df <- subset(df, dem_gov!=.5)
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/dummy.pdf",
height=8, width=8)
boxplot(cig_tax~dem_gov, data=df, names=c("Rep Gov", "Dem Gov"), ylab="Cig Tax")
dev.off()
mod <- lm(cig_tax~dem_gov, data=df)
summary(mod)
nobs(mod)
with(df, by(cig_tax, dem_gov, mean))
with(df, by(cig_tax, dem_gov, sd))
t.test(cig_tax~dem_gov, data=df)
full_df <- read.csv("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/COSP_Limited.csv")
tmp <- full_df[,c("z_cigarette_taxes", "pctfemaleleg",
"govparty_a", "hs_dem_prop_all",
"sen_dem_prop_all", "healthspendpc",
"liberal", "conservative", "state", "year",
"democrat", "z_education_expenditures_per_pup", "incomepcap")]
tmp <- tmp[tmp$year > 2000,]
# tmp <- tmp[tmp$state!="Delaware",]
names(tmp)[1:5] <- c("cig_taxes", "female_leg", "dem_gov",
"house_dem", "sen_dem")
tmp$house_dem <- 1*(tmp$house_dem > .5)
tmp$sen_dem <- 1*(tmp$sen_dem > .5)
mod <- lm(cig_taxes~house_dem, data=tmp)
summary(mod)
tmp$control <- tmp$house_dem + tmp$sen_dem + tmp$dem_gov
tmp$control <- ifelse(tmp$control==0, "Republican",
ifelse(tmp$control==3, "Democrat", "Mixed"))
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/dummy2.pdf",
height=8, width=8)
boxplot(cig_taxes~control, data=tmp, ylab="Cigarette Tax")
dev.off()
mod <- lm(cig_taxes~control, data=tmp)
summary(mod)
nobs(mod)
tmp$control <- as.factor(tmp$control)
mod <- lm(cig_taxes~relevel(control, ref = "Mixed"), data=tmp)
summary(mod)
nobs(mod)
mod <- lm(cig_taxes~relevel(control, ref = "Republican"), data=tmp)
summary(mod)
nobs(mod)
mod <- lm(z_education_expenditures_per_pup~control + incomepcap + liberal , data=tmp)
summary(mod)
nobs(mod)
full_df <- read.csv("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/COSP_Limited.csv")
tmp <- full_df[,c("unemployment",
"govparty_a", "hs_dem_prop_all",
"sen_dem_prop_all",
"state", "year", "labor_minwage_abovefed",
"z_education_expenditures_per_pup",
"welfare_spending_percap",
"incomepcap")]
tmp <- tmp[tmp$year > 1995,]
# tmp <- tmp[tmp$state!="Delaware",]
names(tmp)[2:4] <- c("dem_gov",
"house_dem", "sen_dem")
tmp$house_dem <- 1*(tmp$house_dem > .5)
tmp$sen_dem <- 1*(tmp$sen_dem > .5)
tmp$control <- tmp$house_dem + tmp$sen_dem + tmp$dem_gov
tmp$control <- ifelse(tmp$control==0, "Republican",
ifelse(tmp$control==3, "Democrat", "Mixed"))
tmp$control <- as.factor(tmp$control)
tmp <- tmp[,c("welfare_spending_percap", "control",
"unemployment", "incomepcap", "state", "year")]
names(tmp) <- c("welf_percap", "gov_control", "unemp_rate",
"inc_percap", "state", "year")
tmp$inc_percap <- tmp$inc_percap/1000
tmp <- tmp[tmp$year<2005,]
tmp <- tmp[order(tmp$year), ]
tmp$south <- 0
tmp$south[tmp$state %in% c("Alabama", "Arkansas", "Florida",
"Georgia", "Kansas", "Kentucky",
"Louisiana", "Mississippi",
"Missouri", "North Carolina", "South Carolina",
"Tennessee", "Texas", "Virginia")] <- 1
mod <- lm(welf_percap~ gov_control + unemp_rate +
(inc_percap) + south,
data=tmp)
summary(mod)
nobs(mod)
plot(mod$residuals)
plot(mod$residuals, mod$fitted.values)
summary(tmp)
write.csv(tmp, "/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/MV_Assignment.csv",
row.names = F)
mod <- lm(welf_percap~ relevel(gov_control, ref="Mixed") +
unemp_rate +
(inc_percap) + south,
data=tmp)
summary(mod)
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/outliers.pdf",
height=4, width=4)
par(mar=c(2,2,2,0)+2)
n <- 20
set.seed(3)
x <- rnorm(n, 0, 5)
y <- .25*x + rnorm(n, 0, .5)
plot(y=y, x=x, ylim=c(-6, 6), xlim=c(-12, 22), xlab="X", ylab="Y", main="Good Data!")
abline(lm(y~x))
y_tmp <- c(y, 5)
x_tmp <- c(x, 0)
plot(y=y_tmp, x=x_tmp, ylim=c(-6, 6), xlim=c(-12, 22), xlab="X", ylab="Y", main="Not so Problematic")
abline(lm(y_tmp~x_tmp))
abline(lm(y~x), lty=2)
y_tmp <- c(y, .25*15)
x_tmp <- c(x, 15)
plot(y=y_tmp, x=x_tmp, ylim=c(-6, 6), xlim=c(-12, 22), xlab="X", ylab="Y", main="Not so Problematic")
abline(lm(y_tmp~x_tmp))
abline(lm(y~x), lty=2)
y_tmp <- c(y, -5)
x_tmp <- c(x, 20)
plot(y=y_tmp, x=x_tmp, ylim=c(-6, 6), xlim=c(-12, 22), xlab="X", ylab="Y", main="Very Problematic")
abline(lm(y_tmp~x_tmp))
abline(lm(y~x), lty=2)
dev.off()
n <- 100
set.seed(3)
x <- runif(n, 20, 10000)
y <- log(1*x) + rnorm(n, 0, .2)
plot(y=(y), x=(x))
abline(lm((y)~x))
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/log.pdf",
height=6, width=7)
plot(y=log(seq(1:100)), x=1:100, type='l', ylab='log(x)', xlab='x', main='log (or ln) Transformation')
dev.off()
tmp <- full_df[,c("unemployment",
"govparty_a", "hs_dem_prop_all",
"sen_dem_prop_all",
"state", "year", "labor_minwage_abovefed",
"z_education_expenditures_per_pup",
"welfare_spending_percap",
"incomepcap", "liberal")]
tmp <- tmp[tmp$year > 1995,]
# tmp <- tmp[tmp$state!="Delaware",]
names(tmp)[2:4] <- c("dem_gov",
"house_dem", "sen_dem")
tmp$house_dem <- 1*(tmp$house_dem > .5)
tmp$sen_dem <- 1*(tmp$sen_dem > .5)
tmp$control <- tmp$house_dem + tmp$sen_dem + tmp$dem_gov
tmp$control <- ifelse(tmp$control==0, "Republican",
ifelse(tmp$control==3, "Democrat", "Mixed"))
tmp$control <- as.factor(tmp$control)
tmp <- tmp[,c("welfare_spending_percap", "control",
"unemployment", "incomepcap", "state", "year",
"z_education_expenditures_per_pup", "liberal")]
names(tmp) <- c("welf_percap", "gov_control", "unemp_rate",
"inc_percap", "state", "year", "ed_spending",
"liberal")
tmp$log_incpercap <- log(tmp$inc_percap)
tmp$inc_percap <- (tmp$inc_percap)/1000
tmp$log_welfpercap <- log(tmp$welf_percap)
mod_normal <- lm(inc_percap~
(welf_percap) ,
data=tmp)
summary(mod_normal)
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/lin_log.pdf",
height=6, width=7)
plot(y=tmp$inc_percap, x=tmp$welf_percap, ylim=c(15,50), ylab="Income per Capita ($1k)",
xlab="Welfare per Capita")
plot(y=tmp$inc_percap, x=tmp$welf_percap, ylim=c(15,50), ylab="Income per Capita ($1k)",
xlab="Welfare per Capita", col='gray')
abline(mod_normal)
mod <- lm(inc_percap~
(log_welfpercap) ,
data=tmp)
summary(mod)
nobs(mod)
plot(y=tmp$inc_percap, x=exp(tmp$log_welfpercap), ylim=c(15,50), ylab="Income per Capita ($1k)",
xlab="Welfare per Capita", col='gray')
x <- seq(log(1),
8, by=.1)
y <- coef(mod)[1] + coef(mod)[2]*x
lines(y=y, x=exp(x))
abline(mod_normal, lty=2)
dev.off()
mod <- lm(inc_percap~
(log_welfpercap) + unemp_rate + gov_control ,
data=tmp)
summary(mod)
summary(lm(inc_percap~
(log_welfpercap), data=mod$model))
mod <- lm(log_welfpercap~
(inc_percap) ,
data=tmp)
summary(mod)
plot(y=tmp$welf_percap, x=(tmp$inc_percap))
x <- seq(min(tmp$inc_percap, na.rm=T),
max(tmp$inc_percap, na.rm=T), by=.1)
y <- coef(mod)[1] + coef(mod)[2]*x
lines(y=exp(y), x=(x))
abline(mod_normal)
# tmp <- tmp[tmp$state!="Delaware",]
tmp$ed_spending <- tmp$ed_spending/1000
mod_normal<- lm(welf_percap~
(ed_spending) ,
data=tmp)
summary(mod_normal)
mod<- lm(log_welfpercap~
(ed_spending) ,
data=tmp)
summary(mod)
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/log_lin.pdf",
height=6, width=7)
plot(y=tmp$welf_percap, x=(tmp$ed_spending), col='black', ylab="Welfare per Capita", xlab="Ed Spending per Pupil ($1k)")
plot(y=tmp$welf_percap, x=(tmp$ed_spending), col='gray', ylab="Welfare per Capita", xlab="Ed Spending per Pupil ($1k)")
abline(mod_normal, lty=1)
plot(y=tmp$welf_percap, x=(tmp$ed_spending), col='gray', ylab="Welfare per Capita", xlab="Ed Spending per Pupil ($1k)")
x <- seq(1, 17, by=.1)
y <- coef(mod)[1] + coef(mod)[2]*x
lines(y=exp(y), x=(x))
abline(mod_normal, lty=2)
dev.off()
mod <- lm(log_welfpercap~
(ed_spending) + unemp_rate + gov_control ,
data=tmp)
summary(mod)
summary(lm(log_welfpercap~
(ed_spending), data=mod$model))
tmp <- full_df[,c("unemployment",
"govparty_a", "hs_dem_prop_all",
"sen_dem_prop_all",
"state", "year", "labor_minwage_abovefed",
"z_education_expenditures_per_pup",
"welfare_spending_percap",
"incomepcap", "liberal")]
# tmp <- tmp[tmp$year > 1995,]
# tmp <- tmp[tmp$state!="Delaware",]
tmp$hs_dem_prop_all <- tmp$hs_dem_prop_all*100
tmp$z_education_expenditures_per_pup <- tmp$z_education_expenditures_per_pup/1000
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/poly.pdf",
height=6, width=7)
plot(tmp$hs_dem_prop_all, tmp$z_education_expenditures_per_pup,xlab="Percentage Dems in House",
ylab="Ed Spending per Pupil ($1k)")
mod_normal <- lm(z_education_expenditures_per_pup~hs_dem_prop_all, data=tmp)
mod <- lm(z_education_expenditures_per_pup~hs_dem_prop_all + I(hs_dem_prop_all^2), data=tmp)
summary(mod)
summary(mod_normal)
plot(tmp$hs_dem_prop_all, tmp$z_education_expenditures_per_pup,xlab="Percentage Dems in House",
ylab="Ed Spending per Pupil ($1k)", col='gray')
abline(mod_normal, lty=1)
plot(tmp$hs_dem_prop_all, tmp$z_education_expenditures_per_pup,xlab="Percentage Dems in House",
ylab="Ed Spending per Pupil ($1k)", col='gray')
x <- seq(10, 100, by=1)
y <- coef(mod)[1] + coef(mod)[2] * x + coef(mod)[3] *( x^2)
lines(y=y, x=x)
abline(mod_normal, lty=2)
dev.off()
plot(full_df$hs_dem_prop_all, full_df$z_tanf_maxpayment)
full_df$hs_dem_prop_all <- full_df$hs_dem_prop_all*100
mod <- lm(z_education_expenditures_per_pup~hs_dem_prop_all + I(hs_dem_prop_all^2) +
log(welfare_spending_percap) + unemployment +
labor_minwage_abovefed + I(100*pop5to17/poptotal), data=full_df)
summary(mod)
new_df <- as.data.frame(t(apply(mod$model, 2, median)))
new_df <- new_df[rep(1, 76),]
new_df$hs_dem_prop_all <- 12:87
new_df$`I(hs_dem_prop_all^2)` <- (12:87)^2
y <- as.matrix(cbind(1, new_df[,-1])) %*% (coef(mod))
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/preds.pdf",
height=6, width=7)
plot(y=y, x=12:87, type='l', ylab="Ed Spending per Pupil ($1k)", xlab="Percent Dems")
new_df <- as.data.frame(t(apply(mod$model, 2, median)))
new_df <- new_df[rep(1, 23),]
new_df$`log(welfare_spending_percap)` <- seq(5.5, 7.7, by=.1)
y <- as.matrix(cbind(1, new_df[,-1])) %*% (coef(mod))
plot(y=y, x=exp(seq(5.5, 7.7, by=.1)), type='l', ylab="Ed Spending per Pupil ($1k)",
xlab="Welfare Spending per Capita")
dev.off()
summary(mod)
nobs(mod)
full_df$south <- 0
full_df$south[full_df$state %in% c("Alabama", "Arkansas", "Florida",
"Georgia", "Kansas", "Kentucky",
"Louisiana", "Mississippi",
"Missouri", "North Carolina", "South Carolina",
"Tennessee", "Texas", "Virginia")] <- 1
tmp_df <- full_df[full_df$year<1990,]
plot(tmp_df$hs_dem_prop_all, tmp_df$z_education_expenditures_per_pup, pch=tmp_df$south+1)
mod <- lm(z_education_expenditures_per_pup~hs_dem_prop_all*south, data=tmp_df)
summary(mod)
library(ggplot2)
library(ggthemes)
tmp_df$south <- factor(tmp_df$south, labels=c("Not-South", "South"))
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/inter.pdf",
height=6, width=7)
ggplot(tmp_df, aes(y=z_education_expenditures_per_pup, x=hs_dem_prop_all)) +
# geom_point(aes(color=south), alpha=.5) +
geom_point( alpha=.5) +
scale_color_calc("South") +
labs(y="Education Spending per Capita", x="Democrats in Lower Leg") +
# facet_wrap(~south) +
theme_minimal() +
# geom_smooth(method='lm', se=F) +
NULL
ggplot(tmp_df, aes(y=z_education_expenditures_per_pup, x=hs_dem_prop_all)) +
# geom_point(aes(color=south), alpha=.5) +
geom_point( alpha=.5) +
scale_color_calc("South") +
labs(y="Education Spending per Capita", x="Democrats in Lower Leg") +
# facet_wrap(~south) +
theme_minimal() +
geom_smooth(method='lm', se=F, color='black') +
NULL
ggplot(tmp_df, aes(y=z_education_expenditures_per_pup, x=hs_dem_prop_all)) +
geom_point(aes(color=south), alpha=.5) +
scale_color_calc("South") +
labs(y="Education Spending per Capita", x="Democrats in Lower Leg") +
# facet_wrap(~south) +
theme_minimal() +
geom_smooth(method='lm', se=F, color='black') +
NULL
# ggplot(tmp_df, aes(y=z_education_expenditures_per_pup, x=hs_dem_prop_all)) +
# geom_point(aes(color=south), alpha=.5) +
# scale_color_calc("South") +
# labs(y="Education Spending per Capita", x="Democrats in Lower Leg") +
# # facet_wrap(~south) +
# theme_minimal() +
# # geom_smooth(method='lm', se=F) +
# NULL
ggplot(tmp_df, aes(y=z_education_expenditures_per_pup, x=hs_dem_prop_all)) +
geom_point(aes(color=south), alpha=.5) +
scale_color_calc("South") +
labs(y="Education Spending per Capita", x="Democrats in Lower Leg") +
facet_wrap(~south) +
theme_minimal() +
geom_smooth(method='lm', se=F, color='black') +
NULL
ggplot(tmp_df, aes(y=z_education_expenditures_per_pup, x=hs_dem_prop_all)) +
geom_point(aes(color=south), alpha=.5) +
scale_color_calc("South") +
labs(y="Education Spending per Capita", x="Democrats in Lower Leg") +
# facet_wrap(~south) +
theme_minimal() +
geom_smooth(method='lm', se=F, aes(color=south)) +
NULL
dev.off()
tmp_df <- full_df[full_df$year>2000,]
tmp_df$gini_coef
mod <- lm(welfare_spending_percap~hs_dem_prop_all*gini_coef, data=tmp_df)
summary(mod)
nobs(mod)
# plot(mod)
range(mod$model$gini_coef)
gini <- seq(.52, .71, by=.01)
coef <- coef(mod)[2] + coef(mod)[4]*gini
se <- sqrt(vcov(mod)[2,2] + gini^2 * vcov(mod)[4,4] + 2* gini* vcov(mod)[2,4])
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/inter_cont.pdf",
height=6, width=7)
plot(y=coef, x=gini, type="l", ylim=c(-5,25), ylab="Effect of % Dems", xlab="Gini")
lines(y=coef + 2*se, x=gini, lty=2)
lines(y=coef - 2*se, x=gini, lty=2)
abline(h=0, col='gray')
dev.off()
mod
range(mod$model$hs_dem_prop_all)
hs <- seq(12, 87, by=1)
coef <- coef(mod)[3] + coef(mod)[4]*hs
se <- sqrt(vcov(mod)[3,3] + hs^2 * vcov(mod)[4,4] + 2* hs* vcov(mod)[3,4])
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/inter_cont2.pdf",
height=6, width=7)
plot(y=coef, x=hs, type="l", ylim=c(-5000,6500), ylab="Effect of Gini", xlab="% Dems")
lines(y=coef + 2*se, x=hs, lty=2)
lines(y=coef - 2*se, x=hs, lty=2)
abline(h=0, col='gray')
dev.off()
reuning/POL306 documentation built on Sept. 19, 2023, 10:47 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
setwd("~/Dropbox/Teaching_Kevin/POL306/class_scripts/")
df <- read.csv("qog_oecd_cs_jan21.csv")
df$gea_ea1524f
df$wdi_fertility
mod <- lm(cpds_vt ~ cpds_enps, data=df)
summary(mod)
png("turnout_v_party.png", height=4, width=6, units="in", res=200)
plot(y=df$cpds_vt, x=df$cpds_enps, xlab="Effective Number of Parties",
ylab="Voter Turnout", main="Turnout vs Number of Parties")
abline(mod)
dev.off()
stargazer::stargazer(mod, se =NA,
covariate.labels = c("Effective Number of Parties", "Intercept"),
dep.var.labels = "Turnout",
no.space = T, omit.stat = c("rsq", "adj.rsq", "f", "ser"), type="html")
df <- read.csv("https://www.kevinreuning.com/assets/state_cig.csv")
mod <- lm(cig_tax~liberal, data=df)
summary(mod)
mod <- lm(cig_tax~female_leg, data=df)
summary(mod)
nobs(mod)
mod <- lm(healthspendpc~female_leg, data=df)
summary(mod)
nobs(mod)
confint(mod, level=0.99)
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/resid.pdf",
height=8, width=8)
par(mfrow=c(2,2))
plot(rnorm(1000), ylab="Residual", main="Good!")
plot(rnorm(1000, sd=seq(.5,3, length.out = 1000)), ylab="Residual", main="Heteroskedasticity")
set.seed(1)
tmp <- rnorm(1000)
tmp2 <- tmp + .25*cumsum(tmp)
mean(tmp2)
plot(tmp2, ylab="Residual",main="Autocorrelation")
dev.off()
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/resid2.pdf",
height=6, width=8)
par(mfrow=c(1,2))
x <- rnorm(1000)
y <- 2*x + rnorm(1000)
mod <- lm(y~x)
plot(y=mod$residuals, x=mod$fitted.values, xlab='Predicts', ylab='Residuals', main='Good')
x <- rnorm(1000)
z <- -x + rnorm(1000, 2)
y <- 2*x + rnorm(1000,sd=exp(x))
mod <- lm(y~x)
plot(y=mod$residuals, x=mod$fitted.values, xlab='Predicts', ylab='Residuals', main='Heteroskedasticity')
dev.off()
mod <- lm(cig_tax~female_leg, data=df)
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/resid3.pdf",
height=6, width=8)
par(mfrow=c(1,2))
plot(mod$residuals, ylab='Residuals')
plot(y=mod$residuals, x=mod$fitted.values, xlab='Predicts', ylab='Residuals')
dev.off()
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/null_dist.pdf",
height=6, width=8)
plot(y=dnorm(seq(-.05,.05,by=.001),
sd=sqrt(diag(vcov(mod)))[2] ),
x=seq(-.05,.05, by=.001), type='l', ylab='Density',
xlab='Coefficient Value',
main='Null Distribution')
abline(v=coef(mod)[2], lty=2)
text(x=coef(mod)[2], y=50, pos=3, labels="Estimated Value",
srt=90)
dev.off()
pdf("C:/Users/reunink/Dropbox/Teaching_Kevin/Spring_2021/POL_306/Lectures/images/crit_vals.pdf",
height=6, width=8)
plot(y=dnorm(seq(-3,3,by=.1)),
x=seq(-3,3, by=.1), type='l',
ylab='Density',
xlab="Value",
main="t-distribution")
plot(y=dnorm(seq(-3,3,by=.1)),
x=seq(-3,3, by=.1), type='l',
ylab='Density',
xlab="Value",
main="t-distribution")
polygon(x=rev(c(-1.644854,seq(-1.644854, -3, by=-.1))),
y=rev(c(0.004992899,dnorm(seq(1.644854, 3, by=.1)))),
col='gray')
abline(v=-1.644854)
arrows(y1=dnorm(2), x=-2,
y0=0.15, x0=-2.25
)
text(x=-2.25, y=.15,
label='0.05 Area', pos=3)
text(x=-1.644854, pos=3, y=.2,
label='-1.644854 Crit Value', xpd=T,
srt=270)
plot(y=dnorm(seq(-3,3,by=.1)),
x=seq(-3,3, by=.1), type='l',
ylab='Density',
xlab="Value",
main="t-distribution")
polygon(x=c(seq(-3, -2.326348, length.out = 20), -2.326348),
y=c(dnorm(seq(-3, -2.326348, length.out = 20)), 0.004992899),
col='gray')
abline(v=-2.326348)
arrows(y1=dnorm(-2.5), x1=-2.5,
y0=0.15, x0=-2.75
)
text(x=-2.75, y=.15,
label='0.01 Area', pos=3)
text(x=-2.326348, pos=3, y=.2,
label='-2.326348 Crit Value', xpd=T,
srt=270)
plot(y=dnorm(seq(-3,3,by=.1)),
x=seq(-3,3, by=.1), type='l',
ylab='Density',
xlab="Value",
main="t-distribution")
polygon(x=c(seq(-3, -1.959964, length.out = 20), -1.959964),
y=c(dnorm(seq(-3, -1.959964, length.out = 20)), 0.004992899),
col='gray')
abline(v=-1.959964)
arrows(y1=dnorm(-2.25), x1=-2.25,
y0=0.15, x0=-2.75
)
text(x=-2.75, y=.15,
label='0.025 Area', pos=3)
text(x=-1.959964, pos=3, y=.2,
label='-1.959964 Crit Value', xpd=T,
srt=270)
polygon(x=c(1.959964,seq(1.959964, 3, by=.1)),
y=c(0.004992899,dnorm(seq(1.959964, 3, by=.1))),
col='gray')
abline(v=1.959964)
arrows(y1=dnorm(2.25), x1=2.25,
y0=0.15, x0=2.75
)
text(x=2.65, y=.15,
label='0.025 Area', pos=3)
text(x=1.959964, pos=3, y=.2,
label='1.959964 Crit Value', xpd=T,
srt=270)
dev.off()
pnorm(1.68)
qnorm(0.025)
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/null_dist2.pdf",
height=6, width=8)
plot(y=dnorm(seq(-.05,.05,by=.001),
sd=sqrt(diag(vcov(mod)))[2] ),
x=seq(-.05,.05, by=.001), type='l', ylab='Density',
xlab='Coefficient Value',
main='Null Distribution')
abline(v=0.005, lty=2)
text(x=0.005, y=50, pos=3, labels="Estimated Value",
srt=90)
dev.off()
df[df$state=="Ohio",]
df[df$state=="Indiana",]
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/corr_biv.pdf",
height=8, width=8)
plot(df$liberal, df$female_leg, xlab="% Liberal Voters", ylab="% Women Leg")
dev.off()
cor(df$liberal, df$female_leg, use="complete.obs")
mod <- lm(cig_tax~female_leg + liberal, data=df)
summary(mod)
nobs(mod)
confint(mod)
mod <- lm(cig_tax~female_leg , data=df)
summary(mod)
nobs(mod)
mod <- lm(cig_tax~liberal , data=df)
summary(mod)
nobs(mod)
mod <- lm(healthspendpc~cig_tax , data=df)
summary(mod)
nobs(mod)
confint(mod)
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/assums_ex.pdf",
height=8, width=12)
par(mfrow=c(1,2))
plot(mod$residuals)
plot(y=mod$residuals, x=mod$fitted.values)
dev.off()
full_df <- read.csv("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/COSP_Limited.csv")
tmp <- full_df[,c("z_cigarette_taxes", "pctfemaleleg",
"liberal", "conservative", "state", "year",
"democrat",
"hou_chamber")]
tmp <- tmp[tmp$year > 2000,]
tmp <- tmp[tmp$state!="Delaware",]
names(tmp)[1:2] <- c("cig_taxes", "female_leg")
mod1 <- lm(cig_taxes~liberal+hou_chamber, data=tmp)
mod2 <- lm(cig_taxes~hou_chamber, data=tmp)
mod3 <- lm(cig_taxes~liberal, data=tmp)
summary(mod1)
stargazer::stargazer(mod1, mod2, mod3)
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/multico.pdf",
height=8, width=8)
plot(tmp$liberal, tmp$hou_chamber, xlab="Percentage Liberals",
ylab="Legislative Ideology")
dev.off()
cor(tmp$liberal, tmp$hou_chamber, use='complete.obs')
df <- subset(df, dem_gov!=.5)
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/dummy.pdf",
height=8, width=8)
boxplot(cig_tax~dem_gov, data=df, names=c("Rep Gov", "Dem Gov"), ylab="Cig Tax")
dev.off()
mod <- lm(cig_tax~dem_gov, data=df)
summary(mod)
nobs(mod)
with(df, by(cig_tax, dem_gov, mean))
with(df, by(cig_tax, dem_gov, sd))
t.test(cig_tax~dem_gov, data=df)
full_df <- read.csv("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/COSP_Limited.csv")
tmp <- full_df[,c("z_cigarette_taxes", "pctfemaleleg",
"govparty_a", "hs_dem_prop_all",
"sen_dem_prop_all", "healthspendpc",
"liberal", "conservative", "state", "year",
"democrat", "z_education_expenditures_per_pup", "incomepcap")]
tmp <- tmp[tmp$year > 2000,]
# tmp <- tmp[tmp$state!="Delaware",]
names(tmp)[1:5] <- c("cig_taxes", "female_leg", "dem_gov",
"house_dem", "sen_dem")
tmp$house_dem <- 1*(tmp$house_dem > .5)
tmp$sen_dem <- 1*(tmp$sen_dem > .5)
mod <- lm(cig_taxes~house_dem, data=tmp)
summary(mod)
tmp$control <- tmp$house_dem + tmp$sen_dem + tmp$dem_gov
tmp$control <- ifelse(tmp$control==0, "Republican",
ifelse(tmp$control==3, "Democrat", "Mixed"))
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/dummy2.pdf",
height=8, width=8)
boxplot(cig_taxes~control, data=tmp, ylab="Cigarette Tax")
dev.off()
mod <- lm(cig_taxes~control, data=tmp)
summary(mod)
nobs(mod)
tmp$control <- as.factor(tmp$control)
mod <- lm(cig_taxes~relevel(control, ref = "Mixed"), data=tmp)
summary(mod)
nobs(mod)
mod <- lm(cig_taxes~relevel(control, ref = "Republican"), data=tmp)
summary(mod)
nobs(mod)
mod <- lm(z_education_expenditures_per_pup~control + incomepcap + liberal , data=tmp)
summary(mod)
nobs(mod)
full_df <- read.csv("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/COSP_Limited.csv")
tmp <- full_df[,c("unemployment",
"govparty_a", "hs_dem_prop_all",
"sen_dem_prop_all",
"state", "year", "labor_minwage_abovefed",
"z_education_expenditures_per_pup",
"welfare_spending_percap",
"incomepcap")]
tmp <- tmp[tmp$year > 1995,]
# tmp <- tmp[tmp$state!="Delaware",]
names(tmp)[2:4] <- c("dem_gov",
"house_dem", "sen_dem")
tmp$house_dem <- 1*(tmp$house_dem > .5)
tmp$sen_dem <- 1*(tmp$sen_dem > .5)
tmp$control <- tmp$house_dem + tmp$sen_dem + tmp$dem_gov
tmp$control <- ifelse(tmp$control==0, "Republican",
ifelse(tmp$control==3, "Democrat", "Mixed"))
tmp$control <- as.factor(tmp$control)
tmp <- tmp[,c("welfare_spending_percap", "control",
"unemployment", "incomepcap", "state", "year")]
names(tmp) <- c("welf_percap", "gov_control", "unemp_rate",
"inc_percap", "state", "year")
tmp$inc_percap <- tmp$inc_percap/1000
tmp <- tmp[tmp$year<2005,]
tmp <- tmp[order(tmp$year), ]
tmp$south <- 0
tmp$south[tmp$state %in% c("Alabama", "Arkansas", "Florida",
"Georgia", "Kansas", "Kentucky",
"Louisiana", "Mississippi",
"Missouri", "North Carolina", "South Carolina",
"Tennessee", "Texas", "Virginia")] <- 1
mod <- lm(welf_percap~ gov_control + unemp_rate +
(inc_percap) + south,
data=tmp)
summary(mod)
nobs(mod)
plot(mod$residuals)
plot(mod$residuals, mod$fitted.values)
summary(tmp)
write.csv(tmp, "/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/MV_Assignment.csv",
row.names = F)
mod <- lm(welf_percap~ relevel(gov_control, ref="Mixed") +
unemp_rate +
(inc_percap) + south,
data=tmp)
summary(mod)
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/outliers.pdf",
height=4, width=4)
par(mar=c(2,2,2,0)+2)
n <- 20
set.seed(3)
x <- rnorm(n, 0, 5)
y <- .25*x + rnorm(n, 0, .5)
plot(y=y, x=x, ylim=c(-6, 6), xlim=c(-12, 22), xlab="X", ylab="Y", main="Good Data!")
abline(lm(y~x))
y_tmp <- c(y, 5)
x_tmp <- c(x, 0)
plot(y=y_tmp, x=x_tmp, ylim=c(-6, 6), xlim=c(-12, 22), xlab="X", ylab="Y", main="Not so Problematic")
abline(lm(y_tmp~x_tmp))
abline(lm(y~x), lty=2)
y_tmp <- c(y, .25*15)
x_tmp <- c(x, 15)
plot(y=y_tmp, x=x_tmp, ylim=c(-6, 6), xlim=c(-12, 22), xlab="X", ylab="Y", main="Not so Problematic")
abline(lm(y_tmp~x_tmp))
abline(lm(y~x), lty=2)
y_tmp <- c(y, -5)
x_tmp <- c(x, 20)
plot(y=y_tmp, x=x_tmp, ylim=c(-6, 6), xlim=c(-12, 22), xlab="X", ylab="Y", main="Very Problematic")
abline(lm(y_tmp~x_tmp))
abline(lm(y~x), lty=2)
dev.off()
n <- 100
set.seed(3)
x <- runif(n, 20, 10000)
y <- log(1*x) + rnorm(n, 0, .2)
plot(y=(y), x=(x))
abline(lm((y)~x))
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/log.pdf",
height=6, width=7)
plot(y=log(seq(1:100)), x=1:100, type='l', ylab='log(x)', xlab='x', main='log (or ln) Transformation')
dev.off()
tmp <- full_df[,c("unemployment",
"govparty_a", "hs_dem_prop_all",
"sen_dem_prop_all",
"state", "year", "labor_minwage_abovefed",
"z_education_expenditures_per_pup",
"welfare_spending_percap",
"incomepcap", "liberal")]
tmp <- tmp[tmp$year > 1995,]
# tmp <- tmp[tmp$state!="Delaware",]
names(tmp)[2:4] <- c("dem_gov",
"house_dem", "sen_dem")
tmp$house_dem <- 1*(tmp$house_dem > .5)
tmp$sen_dem <- 1*(tmp$sen_dem > .5)
tmp$control <- tmp$house_dem + tmp$sen_dem + tmp$dem_gov
tmp$control <- ifelse(tmp$control==0, "Republican",
ifelse(tmp$control==3, "Democrat", "Mixed"))
tmp$control <- as.factor(tmp$control)
tmp <- tmp[,c("welfare_spending_percap", "control",
"unemployment", "incomepcap", "state", "year",
"z_education_expenditures_per_pup", "liberal")]
names(tmp) <- c("welf_percap", "gov_control", "unemp_rate",
"inc_percap", "state", "year", "ed_spending",
"liberal")
tmp$log_incpercap <- log(tmp$inc_percap)
tmp$inc_percap <- (tmp$inc_percap)/1000
tmp$log_welfpercap <- log(tmp$welf_percap)
mod_normal <- lm(inc_percap~
(welf_percap) ,
data=tmp)
summary(mod_normal)
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/lin_log.pdf",
height=6, width=7)
plot(y=tmp$inc_percap, x=tmp$welf_percap, ylim=c(15,50), ylab="Income per Capita ($1k)",
xlab="Welfare per Capita")
plot(y=tmp$inc_percap, x=tmp$welf_percap, ylim=c(15,50), ylab="Income per Capita ($1k)",
xlab="Welfare per Capita", col='gray')
abline(mod_normal)
mod <- lm(inc_percap~
(log_welfpercap) ,
data=tmp)
summary(mod)
nobs(mod)
plot(y=tmp$inc_percap, x=exp(tmp$log_welfpercap), ylim=c(15,50), ylab="Income per Capita ($1k)",
xlab="Welfare per Capita", col='gray')
x <- seq(log(1),
8, by=.1)
y <- coef(mod)[1] + coef(mod)[2]*x
lines(y=y, x=exp(x))
abline(mod_normal, lty=2)
dev.off()
mod <- lm(inc_percap~
(log_welfpercap) + unemp_rate + gov_control ,
data=tmp)
summary(mod)
summary(lm(inc_percap~
(log_welfpercap), data=mod$model))
mod <- lm(log_welfpercap~
(inc_percap) ,
data=tmp)
summary(mod)
plot(y=tmp$welf_percap, x=(tmp$inc_percap))
x <- seq(min(tmp$inc_percap, na.rm=T),
max(tmp$inc_percap, na.rm=T), by=.1)
y <- coef(mod)[1] + coef(mod)[2]*x
lines(y=exp(y), x=(x))
abline(mod_normal)
# tmp <- tmp[tmp$state!="Delaware",]
tmp$ed_spending <- tmp$ed_spending/1000
mod_normal<- lm(welf_percap~
(ed_spending) ,
data=tmp)
summary(mod_normal)
mod<- lm(log_welfpercap~
(ed_spending) ,
data=tmp)
summary(mod)
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/log_lin.pdf",
height=6, width=7)
plot(y=tmp$welf_percap, x=(tmp$ed_spending), col='black', ylab="Welfare per Capita", xlab="Ed Spending per Pupil ($1k)")
plot(y=tmp$welf_percap, x=(tmp$ed_spending), col='gray', ylab="Welfare per Capita", xlab="Ed Spending per Pupil ($1k)")
abline(mod_normal, lty=1)
plot(y=tmp$welf_percap, x=(tmp$ed_spending), col='gray', ylab="Welfare per Capita", xlab="Ed Spending per Pupil ($1k)")
x <- seq(1, 17, by=.1)
y <- coef(mod)[1] + coef(mod)[2]*x
lines(y=exp(y), x=(x))
abline(mod_normal, lty=2)
dev.off()
mod <- lm(log_welfpercap~
(ed_spending) + unemp_rate + gov_control ,
data=tmp)
summary(mod)
summary(lm(log_welfpercap~
(ed_spending), data=mod$model))
tmp <- full_df[,c("unemployment",
"govparty_a", "hs_dem_prop_all",
"sen_dem_prop_all",
"state", "year", "labor_minwage_abovefed",
"z_education_expenditures_per_pup",
"welfare_spending_percap",
"incomepcap", "liberal")]
# tmp <- tmp[tmp$year > 1995,]
# tmp <- tmp[tmp$state!="Delaware",]
tmp$hs_dem_prop_all <- tmp$hs_dem_prop_all*100
tmp$z_education_expenditures_per_pup <- tmp$z_education_expenditures_per_pup/1000
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/poly.pdf",
height=6, width=7)
plot(tmp$hs_dem_prop_all, tmp$z_education_expenditures_per_pup,xlab="Percentage Dems in House",
ylab="Ed Spending per Pupil ($1k)")
mod_normal <- lm(z_education_expenditures_per_pup~hs_dem_prop_all, data=tmp)
mod <- lm(z_education_expenditures_per_pup~hs_dem_prop_all + I(hs_dem_prop_all^2), data=tmp)
summary(mod)
summary(mod_normal)
plot(tmp$hs_dem_prop_all, tmp$z_education_expenditures_per_pup,xlab="Percentage Dems in House",
ylab="Ed Spending per Pupil ($1k)", col='gray')
abline(mod_normal, lty=1)
plot(tmp$hs_dem_prop_all, tmp$z_education_expenditures_per_pup,xlab="Percentage Dems in House",
ylab="Ed Spending per Pupil ($1k)", col='gray')
x <- seq(10, 100, by=1)
y <- coef(mod)[1] + coef(mod)[2] * x + coef(mod)[3] *( x^2)
lines(y=y, x=x)
abline(mod_normal, lty=2)
dev.off()
plot(full_df$hs_dem_prop_all, full_df$z_tanf_maxpayment)
full_df$hs_dem_prop_all <- full_df$hs_dem_prop_all*100
mod <- lm(z_education_expenditures_per_pup~hs_dem_prop_all + I(hs_dem_prop_all^2) +
log(welfare_spending_percap) + unemployment +
labor_minwage_abovefed + I(100*pop5to17/poptotal), data=full_df)
summary(mod)
new_df <- as.data.frame(t(apply(mod$model, 2, median)))
new_df <- new_df[rep(1, 76),]
new_df$hs_dem_prop_all <- 12:87
new_df$`I(hs_dem_prop_all^2)` <- (12:87)^2
y <- as.matrix(cbind(1, new_df[,-1])) %*% (coef(mod))
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/preds.pdf",
height=6, width=7)
plot(y=y, x=12:87, type='l', ylab="Ed Spending per Pupil ($1k)", xlab="Percent Dems")
new_df <- as.data.frame(t(apply(mod$model, 2, median)))
new_df <- new_df[rep(1, 23),]
new_df$`log(welfare_spending_percap)` <- seq(5.5, 7.7, by=.1)
y <- as.matrix(cbind(1, new_df[,-1])) %*% (coef(mod))
plot(y=y, x=exp(seq(5.5, 7.7, by=.1)), type='l', ylab="Ed Spending per Pupil ($1k)",
xlab="Welfare Spending per Capita")
dev.off()
summary(mod)
nobs(mod)
full_df$south <- 0
full_df$south[full_df$state %in% c("Alabama", "Arkansas", "Florida",
"Georgia", "Kansas", "Kentucky",
"Louisiana", "Mississippi",
"Missouri", "North Carolina", "South Carolina",
"Tennessee", "Texas", "Virginia")] <- 1
tmp_df <- full_df[full_df$year<1990,]
plot(tmp_df$hs_dem_prop_all, tmp_df$z_education_expenditures_per_pup, pch=tmp_df$south+1)
mod <- lm(z_education_expenditures_per_pup~hs_dem_prop_all*south, data=tmp_df)
summary(mod)
library(ggplot2)
library(ggthemes)
tmp_df$south <- factor(tmp_df$south, labels=c("Not-South", "South"))
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/inter.pdf",
height=6, width=7)
ggplot(tmp_df, aes(y=z_education_expenditures_per_pup, x=hs_dem_prop_all)) +
# geom_point(aes(color=south), alpha=.5) +
geom_point( alpha=.5) +
scale_color_calc("South") +
labs(y="Education Spending per Capita", x="Democrats in Lower Leg") +
# facet_wrap(~south) +
theme_minimal() +
# geom_smooth(method='lm', se=F) +
NULL
ggplot(tmp_df, aes(y=z_education_expenditures_per_pup, x=hs_dem_prop_all)) +
# geom_point(aes(color=south), alpha=.5) +
geom_point( alpha=.5) +
scale_color_calc("South") +
labs(y="Education Spending per Capita", x="Democrats in Lower Leg") +
# facet_wrap(~south) +
theme_minimal() +
geom_smooth(method='lm', se=F, color='black') +
NULL
ggplot(tmp_df, aes(y=z_education_expenditures_per_pup, x=hs_dem_prop_all)) +
geom_point(aes(color=south), alpha=.5) +
scale_color_calc("South") +
labs(y="Education Spending per Capita", x="Democrats in Lower Leg") +
# facet_wrap(~south) +
theme_minimal() +
geom_smooth(method='lm', se=F, color='black') +
NULL
# ggplot(tmp_df, aes(y=z_education_expenditures_per_pup, x=hs_dem_prop_all)) +
# geom_point(aes(color=south), alpha=.5) +
# scale_color_calc("South") +
# labs(y="Education Spending per Capita", x="Democrats in Lower Leg") +
# # facet_wrap(~south) +
# theme_minimal() +
# # geom_smooth(method='lm', se=F) +
# NULL
ggplot(tmp_df, aes(y=z_education_expenditures_per_pup, x=hs_dem_prop_all)) +
geom_point(aes(color=south), alpha=.5) +
scale_color_calc("South") +
labs(y="Education Spending per Capita", x="Democrats in Lower Leg") +
facet_wrap(~south) +
theme_minimal() +
geom_smooth(method='lm', se=F, color='black') +
NULL
ggplot(tmp_df, aes(y=z_education_expenditures_per_pup, x=hs_dem_prop_all)) +
geom_point(aes(color=south), alpha=.5) +
scale_color_calc("South") +
labs(y="Education Spending per Capita", x="Democrats in Lower Leg") +
# facet_wrap(~south) +
theme_minimal() +
geom_smooth(method='lm', se=F, aes(color=south)) +
NULL
dev.off()
tmp_df <- full_df[full_df$year>2000,]
tmp_df$gini_coef
mod <- lm(welfare_spending_percap~hs_dem_prop_all*gini_coef, data=tmp_df)
summary(mod)
nobs(mod)
# plot(mod)
range(mod$model$gini_coef)
gini <- seq(.52, .71, by=.01)
coef <- coef(mod)[2] + coef(mod)[4]*gini
se <- sqrt(vcov(mod)[2,2] + gini^2 * vcov(mod)[4,4] + 2* gini* vcov(mod)[2,4])
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/inter_cont.pdf",
height=6, width=7)
plot(y=coef, x=gini, type="l", ylim=c(-5,25), ylab="Effect of % Dems", xlab="Gini")
lines(y=coef + 2*se, x=gini, lty=2)
lines(y=coef - 2*se, x=gini, lty=2)
abline(h=0, col='gray')
dev.off()
mod
range(mod$model$hs_dem_prop_all)
hs <- seq(12, 87, by=1)
coef <- coef(mod)[3] + coef(mod)[4]*hs
se <- sqrt(vcov(mod)[3,3] + hs^2 * vcov(mod)[4,4] + 2* hs* vcov(mod)[3,4])
pdf("/Users/kevinreuning/Dropbox/Teaching_Kevin/Fall_2019/POL_306/Lectures/images/inter_cont2.pdf",
height=6, width=7)
plot(y=coef, x=hs, type="l", ylim=c(-5000,6500), ylab="Effect of Gini", xlab="% Dems")
lines(y=coef + 2*se, x=hs, lty=2)
lines(y=coef - 2*se, x=hs, lty=2)
abline(h=0, col='gray')
dev.off()
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.