R-Studio中OLS多元回归模型系数的绘图比较
任何人都可以帮我处理我的代码吗？我试图通过模型显示标记颜色的系数的比较。以下是我的代码。学校项目今晚到期。我希望在不同颜色的相同图上比较模型系数。我已搜索过，但无法找到确切的答案，以使我的代码正常工作。我有两个模型，我已合并。合并后，我使用GGPLOT。它绘制系数，但只绘制一个，不按模型比较或颜色代码。
murderdeathModel <-
lm(murder~factor(death_pen)+gini+black_AA+white+poverty,DeathPen1)
summary(murderdeathModel)
MURDERDEATHOUTLIERSREMOVEDMODEL <-
lm(murder~factor(death_pen)+gini+black_AA+white+poverty,DeathPen1)
summary(MURDERDEATHOUTLIERSREMOVEDMODEL)
#TIBBLE of Coef
co1<- summary(murderdeathModel)$coefficients
co1
murderdeathModel <- tibble(est =co1 [,1], # first column of coeftest object
se = co1[,2], # second col.
vars = row.names(co1),
modelname = "mdm") # variable names are the row
names of the matrix
murderdeathModel
MURDERDEATHOUTLIERSREMOVEDMODEL <- tibble(est =co1 [,1], # first column
of coeftest object
se = co1[,2], # second col.
vars = row.names(co1),
modelname = "mdmo") # variable
names are the row names of the matrix
MURDERDEATHOUTLIERSREMOVEDMODEL
#calculate CIs
murderdeathModel <- murderdeathModel %>%
mutate(lb=est-se*1.96,
ub=est+se*1.96) %>%
slice(-1) # get rid of intercept since we generally do not plot
MURDERDEATHOUTLIERSREMOVEDMODEL <- MURDERDEATHOUTLIERSREMOVEDMODEL %>%
mutate(lb=est-se*1.96,
ub=est+se*1.96) %>%
slice(-1) # get rid of intercept since we generally do not plot
MURDERDEATHOUTLIERSREMOVEDMODEL
#binding
comparison1 <-
bind_rows(murderdeathModel,MURDERDEATHOUTLIERSREMOVEDMODEL)
comparison1
#plot
comparison1plot <- comparison1 %>%
ggplot (aes(x=est, xmin=lb, xmax=ub, y=reorder(vars, -est), label=
modelname))
comparison1plot
comparison1plot <- comparison1plot + geom_errorbarh(height=0.2) +
geom_point()
comparison1plot
comparison1plot <- comparison1plot + geom_vline(aes(xintercept=0), color=
"red" , linetype=2) + labs(colour="modelname")
comparison1plot
comparison1plot <- comparison1plot + labs(y="", x="Estimates with 95%
confidence intervals",
title= "Comparison of
Coefficients of Models",
"Dependent variable: Rate of
Murder")
comparison1plot
以下是我的结果：我没有得到任何错误，但我只为每个变量得到一行系数。我需要一个用于另一种颜色的每个模型。感谢您的反馈到目前为止。
> #3.4 problem set 2 extra credit
>
>
> murderdeathModel <- lm(murder~factor(death_pen)+gini+black_AA+white+poverty,DeathPen1)
> summary(murderdeathModel)
Call:
lm(formula = murder ~ factor(death_pen) + gini + black_AA + white +
poverty, data = DeathPen1)
Residuals:
Min 1Q Median 3Q Max
-3.3764 -1.2685 -0.2335 1.1873 4.0976
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 15.30933 8.22860 1.861 0.0695 .
factor(death_pen)death_penalty -0.14468 0.59787 -0.242 0.8099
gini -0.34977 0.18169 -1.925 0.0607 .
black_AA 0.21333 0.03356 6.356 1.01e-07 ***
white -0.04582 0.01882 -2.435 0.0190 *
poverty 0.54859 0.11498 4.771 2.04e-05 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
Residual standard error: 1.836 on 44 degrees of freedom
Multiple R-squared: 0.7335, Adjusted R-squared: 0.7032
F-statistic: 24.22 on 5 and 44 DF, p-value: 1.262e-11
>
> MURDERDEATHOUTLIERSREMOVEDMODEL <- lm(murder~factor(death_pen)+gini+black_AA+white+poverty,DeathPen1)
> summary(MURDERDEATHOUTLIERSREMOVEDMODEL)
Call:
lm(formula = murder ~ factor(death_pen) + gini + black_AA + white +
poverty, data = DeathPen1)
Residuals:
Min 1Q Median 3Q Max
-3.3764 -1.2685 -0.2335 1.1873 4.0976
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 15.30933 8.22860 1.861 0.0695 .
factor(death_pen)death_penalty -0.14468 0.59787 -0.242 0.8099
gini -0.34977 0.18169 -1.925 0.0607 .
black_AA 0.21333 0.03356 6.356 1.01e-07 ***
white -0.04582 0.01882 -2.435 0.0190 *
poverty 0.54859 0.11498 4.771 2.04e-05 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
Residual standard error: 1.836 on 44 degrees of freedom
Multiple R-squared: 0.7335, Adjusted R-squared: 0.7032
F-statistic: 24.22 on 5 and 44 DF, p-value: 1.262e-11
>
> #TIBBLE of Coef
> co1<- summary(murderdeathModel)$coefficients
> co1
Estimate Std. Error t value Pr(>|t|)
(Intercept) 15.30932648 8.22860074 1.860502 6.950603e-02
factor(death_pen)death_penalty -0.14468497 0.59786688 -0.242002 8.099030e-01
gini -0.34976695 0.18169153 -1.925059 6.070169e-02
black_AA 0.21333409 0.03356404 6.356032 1.010401e-07
white -0.04581836 0.01881759 -2.434869 1.902077e-02
poverty 0.54859221 0.11497922 4.771229 2.044677e-05
>
> murderdeathModel <- tibble(est =co1 [,1], # first column of coeftest object
+ se = co1[,2], # second col.
+ vars = row.names(co1),
+ modelname = "mdm") # variable names are the row names of the matrix
> murderdeathModel
# A tibble: 6 x 4
est se vars modelname
<dbl> <dbl> <chr> <chr>
1 15.3 8.23 (Intercept) mdm
2 - 0.145 0.598 factor(death_pen)death_penalty mdm
3 - 0.350 0.182 gini mdm
4 0.213 0.0336 black_AA mdm
5 - 0.0458 0.0188 white mdm
6 0.549 0.115 poverty mdm
> MURDERDEATHOUTLIERSREMOVEDMODEL <- tibble(est =co1 [,1], # first column of coeftest object
+ se = co1[,2], # second col.
+ vars = row.names(co1),
+ modelname = "mdmo") # variable names are the row names of the matrix
> MURDERDEATHOUTLIERSREMOVEDMODEL
# A tibble: 6 x 4
est se vars modelname
<dbl> <dbl> <chr> <chr>
1 15.3 8.23 (Intercept) mdmo
2 - 0.145 0.598 factor(death_pen)death_penalty mdmo
3 - 0.350 0.182 gini mdmo
4 0.213 0.0336 black_AA mdmo
5 - 0.0458 0.0188 white mdmo
6 0.549 0.115 poverty mdmo
> #calculate CIs
> murderdeathModel <- murderdeathModel %>%
+ mutate(lb=est-se*1.96,
+ ub=est+se*1.96) %>%
+ slice(-1) # get rid of intercept since we generally do not plot
> murderdeathModel
# A tibble: 5 x 6
est se vars modelname lb ub
<dbl> <dbl> <chr> <chr> <dbl> <dbl>
1 -0.145 0.598 factor(death_pen)death_penalty mdm -1.32 1.03
2 -0.350 0.182 gini mdm -0.706 0.00635
3 0.213 0.0336 black_AA mdm 0.148 0.279
4 -0.0458 0.0188 white mdm -0.0827 -0.00894
5 0.549 0.115 poverty mdm 0.323 0.774
> MURDERDEATHOUTLIERSREMOVEDMODEL <- MURDERDEATHOUTLIERSREMOVEDMODEL %>%
+ mutate(lb=est-se*1.96,
+ ub=est+se*1.96) %>%
+ slice(-1) # get rid of intercept since we generally do not plot
> MURDERDEATHOUTLIERSREMOVEDMODEL
# A tibble: 5 x 6
est se vars modelname lb ub
<dbl> <dbl> <chr> <chr> <dbl> <dbl>
1 -0.145 0.598 factor(death_pen)death_penalty mdmo -1.32 1.03
2 -0.350 0.182 gini mdmo -0.706 0.00635
3 0.213 0.0336 black_AA mdmo 0.148 0.279
4 -0.0458 0.0188 white mdmo -0.0827 -0.00894
5 0.549 0.115 poverty mdmo 0.323 0.774
>
> ##binding
> comparison1 <-
+ bind_rows(murderdeathModel,MURDERDEATHOUTLIERSREMOVEDMODEL)
> comparison1
# A tibble: 10 x 6
est se vars modelname lb ub
<dbl> <dbl> <chr> <chr> <dbl> <dbl>
1 -0.145 0.598 factor(death_pen)death_penalty mdm -1.32 1.03
2 -0.350 0.182 gini mdm -0.706 0.00635
3 0.213 0.0336 black_AA mdm 0.148 0.279
4 -0.0458 0.0188 white mdm -0.0827 -0.00894
5 0.549 0.115 poverty mdm 0.323 0.774
6 -0.145 0.598 factor(death_pen)death_penalty mdmo -1.32 1.03
7 -0.350 0.182 gini mdmo -0.706 0.00635
8 0.213 0.0336 black_AA mdmo 0.148 0.279
9 -0.0458 0.0188 white mdmo -0.0827 -0.00894
10 0.549 0.115 poverty mdmo 0.323 0.774
> #CI'S
> comparison1 <- comparison1 %>%
+ mutate(lb=est-se*1.96,
+ ub=est+se*1.96) %>%
+ slice(-1) # get rid of intercept since we generally do not plot
> comparison1
# A tibble: 9 x 6
est se vars modelname lb ub
<dbl> <dbl> <chr> <chr> <dbl> <dbl>
1 -0.350 0.182 gini mdm -0.706 0.00635
2 0.213 0.0336 black_AA mdm 0.148 0.279
3 -0.0458 0.0188 white mdm -0.0827 -0.00894
4 0.549 0.115 poverty mdm 0.323 0.774
5 -0.145 0.598 factor(death_pen)death_penalty mdmo -1.32 1.03
6 -0.350 0.182 gini mdmo -0.706 0.00635
7 0.213 0.0336 black_AA mdmo 0.148 0.279
8 -0.0458 0.0188 white mdmo -0.0827 -0.00894
9 0.549 0.115 poverty mdmo 0.323 0.774
> #TIBBLE OF COEFS
>
> murderdeathModel <- tibble(est =co1 [,1], # first column of coeftest object
+ se = co1[,2], # second col.
+ vars = row.names(co1),
+ modelname = "mdm") # variable names are the row names of the matrix
> murderdeathModel
# A tibble: 6 x 4
est se vars modelname
<dbl> <dbl> <chr> <chr>
1 15.3 8.23 (Intercept) mdm
2 - 0.145 0.598 factor(death_pen)death_penalty mdm
3 - 0.350 0.182 gini mdm
4 0.213 0.0336 black_AA mdm
5 - 0.0458 0.0188 white mdm
6 0.549 0.115 poverty mdm
> #plot
>
> #plot
> comparison1plot <- comparison1 %>%
+ ggplot (aes(x=est, xmin=lb, xmax=ub, y=reorder(vars, -est), label= modelname))
> comparison1plot
> comparison1plot <- comparison1plot + geom_errorbarh(height=0.2) + geom_point()
> comparison1plot
> comparison1plot <- comparison1plot + geom_vline(aes(xintercept=0), color= "red" , linetype=2) + labs(colour="modelname")
> comparison1plot
> comparison1plot <- comparison1plot + labs(y="", x="Estimates with 95% confidence intervals",
+ title= "Comparison of Coefficients of Models",
+ "Dependent variable: Rate of Murder")
>
> comparison1plot
下面是我的图表 the graph from the code above
我已将modelName重新编码为一个因子 - 以下是我的代码和结果
comparison1plot <- comparison1plot %>%
mutate(modelName = factor(modelName, levels =0:1, labels =c("MDM","MDMO")))
comparison1plot
结果：
> #plot
> comparison1plot <- comparison1plot %>%
+ mutate(modelName = factor(modelName, levels =0:1, labels =c("MDM","MDMO")))
Error in UseMethod("mutate_") :
no applicable method for 'mutate_' applied to an object of class "c('gg', 'ggplot')"
> comparison1plot
我没有更改任何其他代码。我在#plot和#plot之间添加了一些代码。它仍然不会改变图形。
1 个答案:
答案 0 :(得分：0)
以下是如何使用不同颜色绘制不同数据点的简单示例：
df = data.frame(x = 1:200, y = c(1:100, -101:-200),group = as.factor(rep(c(1,2),each=100)))
ggplot(df,aes(x,y,color=group))+geom_line()+theme_classic()
您需要为每个数据点（我的示例中的组列）添加一个标签，作为不同的因子级别。然后使用aes（）中的factor标签指定颜色。
相关问题
最新问题
- 我写了这段代码，但我无法理解我的错误
- 我无法从一个代码实例的列表中删除 None 值，但我可以在另一个实例中。为什么它适用于一个细分市场而不适用于另一个细分市场？
- 是否有可能使 loadstring 不可能等于打印？卢阿
- java中的random.expovariate()
- Appscript 通过会议在 Google 日历中发送电子邮件和创建活动
- 为什么我的 Onclick 箭头功能在 React 中不起作用？
- 在此代码中是否有使用“this”的替代方法？
- 在 SQL Server 和 PostgreSQL 上查询，我如何从第一个表获得第二个表的可视化
- 每千个数字得到
- 更新了城市边界 KML 文件的来源？