To reproduce this document, you have to install R package ggiraphExtra from github.
install.packages("devtools")
devtools::install_github("cardiomoon/ggiraphExtra")This documnet is the vignette part 1. You can find the vignette part 2 at:
http://rpubs.com/cardiomoon/284985
require(ggplot2)
require(plyr)
require(reshape2)
require(ggiraph)
require(rgl)
require(ggiraphExtra)
require(moonBook) # for use of data radialThe radial data contains demographic data and laboratory data of 115 patients performing IVUS(intravascular ultrasound) examination of a radial artery after tansradial coronary angiography. The NTAV(normalized total atheroma volume measured by intravascular ultrasound(IVUS) in cubic mm) is a quantitative measurement of atherosclerosis. Suppose you want to predict the amount of atherosclerosis(NTAV) from age and status of smoking.
fit=lm(NTAV~age+smoking,data=radial)
summary(fit)
Call:
lm(formula = NTAV ~ age + smoking, data = radial)
Residuals:
Min 1Q Median 3Q Max
-52.799 -14.154 -3.824 9.303 91.393
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 26.5082 17.9231 1.479 0.1420
age 0.5515 0.2323 2.374 0.0193 *
smokingnon-smoker 3.8101 8.7821 0.434 0.6652
smokingsmoker 15.8976 9.4225 1.687 0.0944 .
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Residual standard error: 23.47 on 111 degrees of freedom
Multiple R-squared: 0.0787, Adjusted R-squared: 0.0538
F-statistic: 3.161 on 3 and 111 DF, p-value: 0.02749You can visualize this model with ggPredict(). This function uses ggiraph::geom_point_interactive() and ggiraph::geom_path_interactive() functions to make a interactive plot. You can identify the points and see the regression equation with your mouse. In ANCOVA model, the slope of regression lines are all the same. You can see three parallel lines in this model.
ggPredict(fit,interactive=TRUE)You can make 3D plot for this model with ggPredict3d() function. This function uses rgl::plot3d() function to make 3d plot. You can use your mouse to manipulate the plot. The default is that if you click and hold with the left mouse button, you can rotate the plot by dragging it. The right mouse button(or the mouse wheel) is used to resize it, and the middle button changes the perspective in the point of view.
ggPredict3d(fit,radius=2)Alternatively, you can make the facetted plot according to categorical variable. Set the overlay parameter FALSE and you can get a facetted plot.
ggPredict3d(fit,radius=2,overlay=FALSE)You can make a HTML file(Write HTML and Javascript to display a scene in a web browser) with writeWebGL() function.
writeWebGL("test.html")You man make a snapshot and save as a *.png file with snapshot3d() function.
snapshot3d("test1.png")You can make model with interaction with same predictor variables.
fit1=lm(NTAV~age*smoking,data=radial)
summary(fit1)
Call:
lm(formula = NTAV ~ age * smoking, data = radial)
Residuals:
Min 1Q Median 3Q Max
-51.60 -15.38 -2.80 8.20 87.05
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 44.46124 58.73248 0.757 0.451
age 0.28890 0.85036 0.340 0.735
smokingnon-smoker 1.06313 62.02355 0.017 0.986
smokingsmoker -29.57296 63.48579 -0.466 0.642
age:smokingnon-smoker 0.02888 0.90286 0.032 0.975
age:smokingsmoker 0.72427 0.93927 0.771 0.442
Residual standard error: 23.47 on 109 degrees of freedom
Multiple R-squared: 0.09553, Adjusted R-squared: 0.05405
F-statistic: 2.303 on 5 and 109 DF, p-value: 0.04956In one-way ANCOVA with interaction model, the slope of regression lines are different. The regression equations according to smoking status are:
For ex-smoker, y=0.29*x+44.46
For non-smoker, y=0.32*x+45.52
For smoker, y=1.01*x+14.89Because the slopes of regression lines are different in one-way ANCOVA with interaction model, you can see the regression lines cross each other.
ggPredict(fit1,interactive=TRUE)You can make 3D plot for this model with ggPredict3d() function.
ggPredict3d(fit1,radius=2)Suppose you want to predict the amount of atherosclerosis(NTAV) from age and body weight and the status of smoking.
fit2=lm(NTAV~age+weight+smoking,data=radial)
summary(fit2)
Call:
lm(formula = NTAV ~ age + weight + smoking, data = radial)
Residuals:
Min 1Q Median 3Q Max
-54.210 -12.984 -3.611 7.997 87.451
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) -29.6078 25.1687 -1.176 0.24199
age 0.7000 0.2292 3.054 0.00283 **
weight 0.7489 0.2442 3.067 0.00272 **
smokingnon-smoker 4.1419 8.4681 0.489 0.62573
smokingsmoker 13.2881 9.1246 1.456 0.14816
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Residual standard error: 22.63 on 110 degrees of freedom
Multiple R-squared: 0.1513, Adjusted R-squared: 0.1204
F-statistic: 4.901 on 4 and 110 DF, p-value: 0.001127In this model, the regression equations according to the smoking status are:
For ex-smoker, NTAV=0.7*age+0.75*weight+-29.61
For non-smoker, NTAV=0.7*age+0.75*weight+-25.47
For smoker,NTAV=0.7*age+0.75*weight+-16.32Because the slopes of regression lines are all the same, you can see the regression planes are all parallel.
ggPredict(fit2,interactive=TRUE)You can make 3D plot for this model with ggPredict3d() function.
ggPredict3d(fit2,radius=2)You can make overlayed plot with the following R code.
ggPredict3d(fit2,radius=2,overlay=TRUE)You can make model with interaction with same predictor variables.
fit3=lm(NTAV~(age+weight)*smoking,data=radial)
summary(fit3)
Call:
lm(formula = NTAV ~ (age + weight) * smoking, data = radial)
Residuals:
Min 1Q Median 3Q Max
-53.515 -13.451 -4.229 7.816 83.795
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) -73.9307 130.3128 -0.567 0.572
age 0.9075 1.0301 0.881 0.380
weight 1.2398 1.2260 1.011 0.314
smokingnon-smoker 72.4766 134.1433 0.540 0.590
smokingsmoker 30.1693 135.4396 0.223 0.824
age:smokingnon-smoker -0.4406 1.0747 -0.410 0.683
age:smokingsmoker 0.1472 1.1012 0.134 0.894
weight:smokingnon-smoker -0.6346 1.2683 -0.500 0.618
weight:smokingsmoker -0.3964 1.2925 -0.307 0.760
Residual standard error: 22.88 on 106 degrees of freedom
Multiple R-squared: 0.1644, Adjusted R-squared: 0.1014
F-statistic: 2.607 on 8 and 106 DF, p-value: 0.0121Because the slopes of regression lines are different, you can see the regression lines are not parallel among groups.
ggPredict(fit3,interactive=TRUE)You can make 3D plot for this model with ggPredict3d() function.
ggPredict3d(fit3,radius=2)You can make overlayed plot with the following R code. In this plot you can see the regression planes cross each other.
ggPredict3d(fit3,radius=2,overlay=TRUE)You can visualize the two-way factorial ANOVA model.
fit4=lm(NTAV~age*DM*smoking,data=radial)
summary(fit4)
Call:
lm(formula = NTAV ~ age * DM * smoking, data = radial)
Residuals:
Min 1Q Median 3Q Max
-52.869 -12.669 -2.989 10.970 80.747
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 3.931e+01 6.283e+01 0.626 0.533
age 3.643e-01 9.439e-01 0.386 0.700
DM 1.357e+02 3.109e+02 0.436 0.663
smokingnon-smoker -5.997e-01 6.686e+01 -0.009 0.993
smokingsmoker -1.198e+01 6.766e+01 -0.177 0.860
age:DM -1.785e+00 4.111e+00 -0.434 0.665
age:smokingnon-smoker -4.112e-03 1.007e+00 -0.004 0.997
age:smokingsmoker 5.599e-01 1.034e+00 0.541 0.589
DM:smokingnon-smoker -1.071e+02 3.138e+02 -0.341 0.733
DM:smokingsmoker -2.679e+02 3.183e+02 -0.842 0.402
age:DM:smokingnon-smoker 1.523e+00 4.160e+00 0.366 0.715
age:DM:smokingsmoker 3.585e+00 4.256e+00 0.842 0.402
Residual standard error: 22.63 on 103 degrees of freedom
Multiple R-squared: 0.2054, Adjusted R-squared: 0.1205
F-statistic: 2.42 on 11 and 103 DF, p-value: 0.01014ggPredict(fit4,interactive=TRUE)You can make 3D plot for this model with ggPredict3d() function.
ggPredict3d(fit4,radius=2)You can make facetted plot with the following R code. In this plot you can see the regression planes cross each other.
ggPredict3d(fit4,radius=2,overlay=TRUE)