R/depbcomp.R
In chrisaberson/BetterReg: Better Statistics for OLS and Binomial Logistic Regression
Defines functions
depbcomp
Documented in
depbcomp
#'Power for Comparing Dependent Coefficients in Multiple Regression with Two or Three Predictors
#'Requires correlations between all variables as sample size. Means, sds, and alpha are option. Also computes Power(All)
#'@param data name of your datafile, loaded
#'@param y dependent variable name
#'@param x1 first predictor variable name
#'@param x2 second predictor variable name
#'@param x3 third predictor variable name
#'@param x4 fourth predictor variable name
#'@param x5 fifth predictor variable name
#'@param data name of data file
#'@param numpred number of predictors
#'@param comps Type of comparison, "abs" for absolute values or "raw" for raw coefficients
#'@examples
#'depbcomp(data=testreg,y="y",x1="x1",x2="x2",x3="x3",x4="x4",x5="x5", numpred=5,comps="abs")
#'@return Comparing Dependent Coefficients in Multiple Regression
#'@export
#'
depbcomp<-function(data=NULL,y=NULL, x1=NULL, x2=NULL,x3=NULL,x4=NULL,x5=NULL,numpred=NULL, comps="abs")
{
if (numpred ==2 && comps=="abs")
{
xx<-dplyr::select(data,y,x1,x2)
xx<-as.data.frame(xx)
mod<-lm(xx[,1]~xx[,2]+xx[,3])
values<-summary(mod)
b1<-(values$coefficients)[2,1] #grabs b from each analysis
b2<-(values$coefficients)[3,1]
seb1<-(values$coefficients)[2,2]
seb2<-(values$coefficients)[3,2]
df<- values$df[2]
varx1<-var(xx[,2])
varx2<-var(xx[,3])
r12<-abs(cor(xx[,2],xx[,3]))
mat<-cbind(c(1,r12),c(r12,1))
inv<-solve(mat)*mat
pij<-inv[1,2] #inv of cor between pred1 and 2
pii<-inv[1,1] #inv of cov, v1
pjj<-inv[2,2] #inv of cov, v2
den1<-seb1^2+seb2^2
den2<-2*seb1*seb2
den3<-pij/(pii+pjj)
den<-(den1-(den2*den3))^.5
t12<-abs((abs(b1)-(abs(b2)))) / den
p12<-2*(1-(pt(q=abs(t12), df=df, lower.tail=TRUE)))
t12<-round((t12),3)
message("Pred 1 vs. Pred 2 "," : t = ", t12,", p = ", p12)
}
if (numpred ==3 && comps=="abs")
{
xx<-dplyr::select(data,y,x1,x2,x3)
xx<-as.data.frame(xx)
mod<-lm(xx[,1]~xx[,2]+xx[,3]+xx[,4])
values<-summary(mod)
b1<-(values$coefficients)[2,1] #grabs b from each analysis
b2<-(values$coefficients)[3,1]
b3<-(values$coefficients)[4,1]
seb1<-(values$coefficients)[2,2]
seb2<-(values$coefficients)[3,2]
seb3<-(values$coefficients)[4,2]
df<- values$df[2]
r12<-abs(cor(xx[,2],xx[,3]))
r13<-abs(cor(xx[,2],xx[,4]))
r23<-abs(cor(xx[,3],xx[,4]))
mat<-cbind(c(1,r12,r13),c(r12,1,r23), c(r13,r23,1))
inv<-solve(mat)*mat
# 1 vs 2
pij1<-inv[1,2] #inv of cor between pred of interest 1 vs. 2
pii1<-inv[1,1] #inv of cov, v1
pjj1<-inv[2,2] #inv of cov, v2
den1a<-seb1^2+seb2^2
den2a<-2*seb1*seb2
den3a<-pij1/(pii1+pjj1)
den<-(den1a-(den2a*den3a))^.5
t12<-abs((abs(b1)-(abs(b2)))) / den
p12<-2*(1-(pt(q=abs(t12), df=df, lower.tail=TRUE)))
t12<-round((t12),3)
#1 vs 3
pij2<-inv[1,3] #inv of cor between pred of interest 1 vs. 2
pii2<-inv[1,1] #inv of cov, v1
pjj2<-inv[3,3] #inv of cov, v2
den1b<-seb1^2+seb3^2
den2b<-2*seb1*seb3
den3b<-pij2/(pii2+pjj2)
denb<-(den1b-(den2b*den3b))^.5
t13<-abs(abs(b1)-abs(b3)) / denb
p13<-2*(1-(pt(q=abs(t13), df=df, lower.tail=TRUE)))
t13<-round((t13),3)
#2 vs 3
pij3<-inv[2,3] #inv of cor between pred of interest 1 vs. 2
pii3<-inv[2,2] #inv of cov, v1
pjj3<-inv[3,3] #inv of cov, v2
den1c<-seb2^2+seb3^2
den2c<-2*seb2*seb3
den3c<-pij3/(pii3+pjj3)
denc<-(den1c-(den2c*den3c))^.5
t23<-abs(abs(b2)-abs(b3)) / denc
p23<-2*(1-(pt(q=abs(t23), df=df, lower.tail=TRUE)))
t23<-round((t23),3)
message("Pred 1 vs. Pred 2 "," : t = ", t12,", p = ", p12)
message("Pred 1 vs. Pred 3 "," : t = ", t13,", p = ", p13)
message("Pred 2 vs. Pred 3 "," : t = ", t23,", p = ", p23)
}
if (numpred ==4 && comps=="abs")
{
xx<-dplyr::select(data,y,x1,x2,x3,x4)
xx<-as.data.frame(xx)
mod<-lm(xx[,1]~xx[,2]+xx[,3]+xx[,4]+xx[,5])
values<-summary(mod)
b1<-(values$coefficients)[2,1] #grabs b from each analysis
b2<-(values$coefficients)[3,1]
b3<-(values$coefficients)[4,1]
b4<-(values$coefficients)[5,1]
seb1<-(values$coefficients)[2,2]
seb2<-(values$coefficients)[3,2]
seb3<-(values$coefficients)[4,2]
seb4<-(values$coefficients)[5,2]
df<- values$df[2]
r12<-abs(cor(xx[,2],xx[,3]))
r13<-abs(cor(xx[,2],xx[,4]))
r14<-abs(cor(xx[,2],xx[,5]))
r23<-abs(cor(xx[,3],xx[,4]))
r24<-abs(cor(xx[,3],xx[,5]))
r34<-abs(cor(xx[,4],xx[,5]))
mat<-cbind(c(1,r12,r13,r14),c(r12,1,r23,r24), c(r13,r23,1,r34),c(r14,r24,r34,1))
inv<-solve(mat)*mat
# 1 vs 2
pij1<-inv[1,2] #inv of cor between pred of interest 1 vs. 2
pii1<-inv[1,1] #inv of cov, v1
pjj1<-inv[2,2] #inv of cov, v2
den1a<-seb1^2+seb2^2
den2a<-2*seb1*seb2
den3a<-pij1/(pii1+pjj1)
den<-(den1a-(den2a*den3a))^.5
t12<-abs((abs(b1)-(abs(b2)))) / den
p12<-2*(1-(pt(q=abs(t12), df=df, lower.tail=TRUE)))
t12<-round((t12),3)
#1 vs 3
pij2<-inv[1,3] #inv of cor between pred of interest 1 vs. 2
pii2<-inv[1,1] #inv of cov, v1
pjj2<-inv[3,3] #inv of cov, v2
den1b<-seb1^2+seb3^2
den2b<-2*seb1*seb3
den3b<-pij2/(pii2+pjj2)
denb<-(den1b-(den2b*den3b))^.5
t13<-abs(abs(b1)-abs(b3)) / denb
p13<-2*(1-(pt(q=abs(t13), df=df, lower.tail=TRUE)))
t13<-round((t13),3)
#1 vs 4
pij4<-inv[1,4] #inv of cor between pred of interest 1 vs. 2
pii4<-inv[1,1] #inv of cov, v1
pjj4<-inv[4,4] #inv of cov, v2
den1d<-seb1^2+seb4^2
den2d<-2*seb1*seb4
den3d<-pij4/(pii4+pjj4)
dend<-(den1d-(den2d*den3d))^.5
t14<-abs(abs(b1)-abs(b4)) / dend
p14<-2*(1-(pt(q=abs(t14), df=df, lower.tail=TRUE)))
t14<-round((t14),3)
#2 vs 3
pij3<-inv[2,3] #inv of cor between pred of interest 1 vs. 2
pii3<-inv[2,2] #inv of cov, v1
pjj3<-inv[3,3] #inv of cov, v2
den1c<-seb2^2+seb3^2
den2c<-2*seb2*seb3
den3c<-pij3/(pii3+pjj3)
denc<-(den1c-(den2c*den3c))^.5
t23<-abs(abs(b2)-abs(b3)) / denc
p23<-2*(1-(pt(q=abs(t23), df=df, lower.tail=TRUE)))
t23<-round((t23),3)
#2 vs 4
pij5<-inv[2,4] #inv of cor between pred of interest 1 vs. 2
pii5<-inv[2,2] #inv of cov, v1
pjj5<-inv[4,4] #inv of cov, v2
den1e<-seb2^2+seb4^2
den2e<-2*seb2*seb4
den3e<-pij5/(pii5+pjj5)
denc<-(den1e-(den2e*den3e))^.5
t24<-abs(abs(b2)-abs(b4)) / denc
p24<-2*(1-(pt(q=abs(t24), df=df, lower.tail=TRUE)))
t24<-round((t24),3)
#3 vs 4
pij6<-inv[3,4] #inv of cor between pred of interest 1 vs. 2
pii6<-inv[3,3] #inv of cov, v1
pjj6<-inv[4,4] #inv of cov, v2
den1f<-seb3^2+seb4^2
den2f<-2*seb3*seb4
den3f<-pij6/(pii6+pjj6)
denc<-(den1f-(den2f*den3f))^.5
t34<-abs(abs(b3)-abs(b4)) / denc
p34<-2*(1-(pt(q=abs(t34), df=df, lower.tail=TRUE)))
t34<-round((t34),3)
message("Pred 1 vs. Pred 2 "," : t = ", t12,", p = ", p12)
message("Pred 1 vs. Pred 3 "," : t = ", t13,", p = ", p13)
message("Pred 1 vs. Pred 4 "," : t = ", t14,", p = ", p14)
message("Pred 2 vs. Pred 3 "," : t = ", t23,", p = ", p23)
message("Pred 2 vs. Pred 4 "," : t = ", t24,", p = ", p24)
message("Pred 3 vs. Pred 4 "," : t = ", t34,", p = ", p34)
}
if (numpred ==5 && comps=="abs")
{
xx<-dplyr::select(data,y,x1,x2,x3,x4,x5)
xx<-as.data.frame(xx)
mod<-lm(xx[,1]~xx[,2]+xx[,3]+xx[,4]+xx[,5]+xx[,6])
values<-summary(mod)
b1<-(values$coefficients)[2,1] #grabs b from each analysis
b2<-(values$coefficients)[3,1]
b3<-(values$coefficients)[4,1]
b4<-(values$coefficients)[5,1]
b5<-(values$coefficients)[6,1]
seb1<-(values$coefficients)[2,2]
seb2<-(values$coefficients)[3,2]
seb3<-(values$coefficients)[4,2]
seb4<-(values$coefficients)[5,2]
seb5<-(values$coefficients)[6,2]
df<- values$df[2]
r12<-abs(cor(xx[,2],xx[,3]))
r13<-abs(cor(xx[,2],xx[,4]))
r14<-abs(cor(xx[,2],xx[,5]))
r15<-abs(cor(xx[,2],xx[,6]))
r23<-abs(cor(xx[,3],xx[,4]))
r24<-abs(cor(xx[,3],xx[,5]))
r25<-abs(cor(xx[,3],xx[,6]))
r34<-abs(cor(xx[,4],xx[,5]))
r35<-abs(cor(xx[,4],xx[,5]))
r45<-abs(cor(xx[,5],xx[,6]))
mat<-cbind(c(1,r12,r13,r14,r15),c(r12,1,r23,r24,r25), c(r13,r23,1,r34,r35),
c(r14,r24,r34,1,r45),c(r15,r25,r35,r45,1))
inv<-solve(mat)*mat
# 1 vs 2
pij1<-inv[1,2] #inv of cor between pred of interest 1 vs. 2
pii1<-inv[1,1] #inv of cov, v1
pjj1<-inv[2,2] #inv of cov, v2
den1a<-seb1^2+seb2^2
den2a<-2*seb1*seb2
den3a<-pij1/(pii1+pjj1)
den<-(den1a-(den2a*den3a))^.5
t12<-abs((abs(b1)-(abs(b2)))) / den
p12<-2*(1-(pt(q=abs(t12), df=df, lower.tail=TRUE)))
t12<-round((t12),3)
#1 vs 3
pij2<-inv[1,3] #inv of cor between pred of interest 1 vs. 2
pii2<-inv[1,1] #inv of cov, v1
pjj2<-inv[3,3] #inv of cov, v2
den1b<-seb1^2+seb3^2
den2b<-2*seb1*seb3
den3b<-pij2/(pii2+pjj2)
denb<-(den1b-(den2b*den3b))^.5
t13<-abs(abs(b1)-abs(b3)) / denb
p13<-2*(1-(pt(q=abs(t13), df=df, lower.tail=TRUE)))
t13<-round((t13),3)
#1 vs 4
pij4<-inv[1,4] #inv of cor between pred of interest 1 vs. 2
pii4<-inv[1,1] #inv of cov, v1
pjj4<-inv[4,4] #inv of cov, v2
den1d<-seb1^2+seb4^2
den2d<-2*seb1*seb4
den3d<-pij4/(pii4+pjj4)
dend<-(den1d-(den2d*den3d))^.5
t14<-abs(abs(b1)-abs(b4)) / dend
p14<-2*(1-(pt(q=abs(t14), df=df, lower.tail=TRUE)))
t14<-round((t14),3)
#1 vs 5
pij6<-inv[1,5] #inv of cor between pred of interest 1 vs. 2
pii6<-inv[1,1] #inv of cov, v1
pjj6<-inv[5,5] #inv of cov, v2
den1f<-seb1^2+seb5^2
den2f<-2*seb1*seb5
den3f<-pij6/(pii4+pjj6)
denf<-(den1f-(den2f*den3f))^.5
t15<-abs(abs(b1)-abs(b5)) / dend
p15<-2*(1-(pt(q=abs(t15), df=df, lower.tail=TRUE)))
t15<-round((t15),3)
#2 vs 3
pij3<-inv[2,3] #inv of cor between pred of interest 1 vs. 2
pii3<-inv[2,2] #inv of cov, v1
pjj3<-inv[3,3] #inv of cov, v2
den1c<-seb2^2+seb3^2
den2c<-2*seb2*seb3
den3c<-pij3/(pii3+pjj3)
denc<-(den1c-(den2c*den3c))^.5
t23<-abs(abs(b2)-abs(b3)) / denc
p23<-2*(1-(pt(q=abs(t23), df=df, lower.tail=TRUE)))
t23<-round((t23),3)
#2 vs 4
pij5<-inv[2,4] #inv of cor between pred of interest 1 vs. 2
pii5<-inv[2,2] #inv of cov, v1
pjj5<-inv[4,4] #inv of cov, v2
den1e<-seb2^2+seb4^2
den2e<-2*seb2*seb4
den3e<-pij5/(pii5+pjj5)
denc<-(den1e-(den2e*den3e))^.5
t24<-abs(abs(b2)-abs(b4)) / denc
p24<-2*(1-(pt(q=abs(t24), df=df, lower.tail=TRUE)))
t24<-round((t24),3)
#2 vs 5
pij7<-inv[2,5] #inv of cor between pred of interest 1 vs. 2
pii7<-inv[2,2] #inv of cov, v1
pjj7<-inv[5,5] #inv of cov, v2
den1g<-seb2^2+seb5^2
den2g<-2*seb2*seb5
den3g<-pij7/(pii7+pjj7)
deng<-(den1g-(den2g*den3g))^.5
t25<-abs(abs(b2)-abs(b5)) / deng
p25<-2*(1-(pt(q=abs(t25), df=df, lower.tail=TRUE)))
t25<-round((t25),3)
#3 vs 4
pij8<-inv[3,4] #inv of cor between pred of interest 1 vs. 2
pii8<-inv[3,3] #inv of cov, v1
pjj8<-inv[4,4] #inv of cov, v2
den1h<-seb3^2+seb4^2
den2h<-2*seb3*seb4
den3h<-pij8/(pii8+pjj8)
denc<-(den1h-(den2h*den3h))^.5
t34<-abs(abs(b3)-abs(b4)) / denc
p34<-2*(1-(pt(q=abs(t34), df=df, lower.tail=TRUE)))
t34<-round((t34),3)
#3 vs 5
pij9<-inv[3,5] #inv of cor between pred of interest 1 vs. 2
pii9<-inv[3,3] #inv of cov, v1
pjj9<-inv[5,5] #inv of cov, v2
den1i<-seb3^2+seb5^2
den2i<-2*seb3*seb5
den3i<-pij9/(pii9+pjj9)
deni<-(den1i-(den2i*den3i))^.5
t35<-abs(abs(b3)-abs(b5)) / denc
p35<-2*(1-(pt(q=abs(t35), df=df, lower.tail=TRUE)))
t35<-round((t35),3)
#4 vs 5
pij10<-inv[4,5] #inv of cor between pred of interest 1 vs. 2
pii10<-inv[4,4] #inv of cov, v1
pjj10<-inv[5,5] #inv of cov, v2
den1j<-seb4^2+seb5^2
den2j<-2*seb4*seb5
den3j<-pij10/(pii10+pjj10)
denj<-(den1h-(den2h*den3h))^.5
t45<-abs(abs(b4)-abs(b5)) / denc
p45<-2*(1-(pt(q=abs(t45), df=df, lower.tail=TRUE)))
t45<-round((t45),3)
message("Pred 1 vs. Pred 2 "," : t = ", t12,", p = ", p12)
message("Pred 1 vs. Pred 3 "," : t = ", t13,", p = ", p13)
message("Pred 1 vs. Pred 4 "," : t = ", t14,", p = ", p14)
message("Pred 1 vs. Pred 5 "," : t = ", t15,", p = ", p15)
message("Pred 2 vs. Pred 3 "," : t = ", t23,", p = ", p23)
message("Pred 2 vs. Pred 4 "," : t = ", t24,", p = ", p24)
message("Pred 2 vs. Pred 5 "," : t = ", t25,", p = ", p25)
message("Pred 3 vs. Pred 4 "," : t = ", t34,", p = ", p34)
message("Pred 3 vs. Pred 5 "," : t = ", t35,", p = ", p35)
message("Pred 4 vs. Pred 5 "," : t = ", t45,", p = ", p45)
}
if (numpred ==2 && comps=="raw")
{
xx<-dplyr::select(data,y,x1,x2)
xx<-as.data.frame(xx)
mod<-lm(xx[,1]~xx[,2]+xx[,3])
values<-summary(mod)
b1<-(values$coefficients)[2,1] #grabs b from each analysis
b2<-(values$coefficients)[3,1]
seb1<-(values$coefficients)[2,2]
seb2<-(values$coefficients)[3,2]
df<- values$df[2]
varx1<-var(xx[,2])
varx2<-var(xx[,3])
r12<-cor(xx[,2],xx[,3])
mat<-cbind(c(1,r12),c(r12,1))
inv<-solve(mat)*mat
pij<-inv[1,2] #inv of cor between pred1 and 2
pii<-inv[1,1] #inv of cov, v1
pjj<-inv[2,2] #inv of cov, v2
den1<-seb1^2+seb2^2
den2<-2*seb1*seb2
den3<-pij/(pii+pjj)
den<-(den1-(den2*den3))^.5
t12<-(b1-b2)/ den
p12<-2*(1-(pt(q=abs(t12), df=df, lower.tail=TRUE)))
t12<-round((t12),3)
message("Pred 1 vs. Pred 2 "," : t = ", t12,", p = ", p12)
}
if (numpred ==3 && comps=="raw")
{
xx<-dplyr::select(data,y,x1,x2,x3)
xx<-as.data.frame(xx)
mod<-lm(xx[,1]~xx[,2]+xx[,3]+xx[,4])
values<-summary(mod)
b1<-(values$coefficients)[2,1] #grabs b from each analysis
b2<-(values$coefficients)[3,1]
b3<-(values$coefficients)[4,1]
seb1<-(values$coefficients)[2,2]
seb2<-(values$coefficients)[3,2]
seb3<-(values$coefficients)[4,2]
df<- values$df[2]
r12<-cor(xx[,2],xx[,3])
r13<-cor(xx[,2],xx[,4])
r23<-cor(xx[,3],xx[,4])
mat<-cbind(c(1,r12,r13),c(r12,1,r23), c(r13,r23,1))
inv<-solve(mat)*mat
# 1 vs 2
pij1<-inv[1,2] #inv of cor between pred of interest 1 vs. 2
pii1<-inv[1,1] #inv of cov, v1
pjj1<-inv[2,2] #inv of cov, v2
den1a<-seb1^2+seb2^2
den2a<-2*seb1*seb2
den3a<-pij1/(pii1+pjj1)
den<-(den1a-(den2a*den3a))^.5
t12<-(b1-b2) / den
p12<-2*(1-(pt(q=abs(t12), df=df, lower.tail=TRUE)))
t12<-round((t12),3)
#1 vs 3
pij2<-inv[1,3] #inv of cor between pred of interest 1 vs. 2
pii2<-inv[1,1] #inv of cov, v1
pjj2<-inv[3,3] #inv of cov, v2
den1b<-seb1^2+seb3^2
den2b<-2*seb1*seb3
den3b<-pij2/(pii2+pjj2)
denb<-(den1b-(den2b*den3b))^.5
t13<-(b1-b3) / denb
p13<-2*(1-(pt(q=abs(t13), df=df, lower.tail=TRUE)))
t13<-round((t13),3)
#2 vs 3
pij3<-inv[2,3] #inv of cor between pred of interest 1 vs. 2
pii3<-inv[2,2] #inv of cov, v1
pjj3<-inv[3,3] #inv of cov, v2
den1c<-seb2^2+seb3^2
den2c<-2*seb2*seb3
den3c<-pij3/(pii3+pjj3)
denc<-(den1c-(den2c*den3c))^.5
t23<-abs(abs(b2)-abs(b3)) / denc
p23<-2*(1-(pt(q=abs(t23), df=df, lower.tail=TRUE)))
t23<-round((t23),3)
message("Pred 1 vs. Pred 2 "," : t = ", t12,", p = ", p12)
message("Pred 1 vs. Pred 3 "," : t = ", t13,", p = ", p13)
message("Pred 2 vs. Pred 3 "," : t = ", t23,", p = ", p23)
}
if (numpred ==4 && comps=="raw")
{
xx<-dplyr::select(data,y,x1,x2,x3,x4)
xx<-as.data.frame(xx)
mod<-lm(xx[,1]~xx[,2]+xx[,3]+xx[,4]+xx[,5])
values<-summary(mod)
b1<-(values$coefficients)[2,1] #grabs b from each analysis
b2<-(values$coefficients)[3,1]
b3<-(values$coefficients)[4,1]
b4<-(values$coefficients)[5,1]
seb1<-(values$coefficients)[2,2]
seb2<-(values$coefficients)[3,2]
seb3<-(values$coefficients)[4,2]
seb4<-(values$coefficients)[5,2]
df<- values$df[2]
r12<-cor(xx[,2],xx[,3])
r13<-cor(xx[,2],xx[,4])
r14<-cor(xx[,2],xx[,5])
r23<-cor(xx[,3],xx[,4])
r24<-cor(xx[,3],xx[,5])
r34<-cor(xx[,4],xx[,5])
mat<-cbind(c(1,r12,r13,r14),c(r12,1,r23,r24), c(r13,r23,1,r34),c(r14,r24,r34,1))
inv<-solve(mat)*mat
# 1 vs 2
pij1<-inv[1,2] #inv of cor between pred of interest 1 vs. 2
pii1<-inv[1,1] #inv of cov, v1
pjj1<-inv[2,2] #inv of cov, v2
den1a<-seb1^2+seb2^2
den2a<-2*seb1*seb2
den3a<-pij1/(pii1+pjj1)
den<-(den1a-(den2a*den3a))^.5
t12<-(b1-b2) / den
p12<-2*(1-(pt(q=abs(t12), df=df, lower.tail=TRUE)))
t12<-round((t12),3)
#1 vs 3
pij2<-inv[1,3] #inv of cor between pred of interest 1 vs. 2
pii2<-inv[1,1] #inv of cov, v1
pjj2<-inv[3,3] #inv of cov, v2
den1b<-seb1^2+seb3^2
den2b<-2*seb1*seb3
den3b<-pij2/(pii2+pjj2)
denb<-(den1b-(den2b*den3b))^.5
t13<-(b1-b3) / denb
p13<-2*(1-(pt(q=abs(t13), df=df, lower.tail=TRUE)))
t13<-round((t13),3)
#1 vs 4
pij4<-inv[1,4] #inv of cor between pred of interest 1 vs. 2
pii4<-inv[1,1] #inv of cov, v1
pjj4<-inv[4,4] #inv of cov, v2
den1d<-seb1^2+seb4^2
den2d<-2*seb1*seb4
den3d<-pij4/(pii4+pjj4)
dend<-(den1d-(den2d*den3d))^.5
t14<-(b1-b4) / dend
p14<-2*(1-(pt(q=abs(t14), df=df, lower.tail=TRUE)))
t14<-round((t14),3)
#2 vs 3
pij3<-inv[2,3] #inv of cor between pred of interest 1 vs. 2
pii3<-inv[2,2] #inv of cov, v1
pjj3<-inv[3,3] #inv of cov, v2
den1c<-seb2^2+seb3^2
den2c<-2*seb2*seb3
den3c<-pij3/(pii3+pjj3)
denc<-(den1c-(den2c*den3c))^.5
t23<-abs(abs(b2)-abs(b3)) / denc
p23<-2*(1-(pt(q=abs(t23), df=df, lower.tail=TRUE)))
t23<-round((t23),3)
#2 vs 4
pij5<-inv[2,4] #inv of cor between pred of interest 1 vs. 2
pii5<-inv[2,2] #inv of cov, v1
pjj5<-inv[4,4] #inv of cov, v2
den1e<-seb2^2+seb4^2
den2e<-2*seb2*seb4
den3e<-pij5/(pii5+pjj5)
denc<-(den1e-(den2e*den3e))^.5
t24<-(b2-b4) / denc
p24<-2*(1-(pt(q=abs(t24), df=df, lower.tail=TRUE)))
t24<-round((t24),3)
#3 vs 4
pij6<-inv[3,4] #inv of cor between pred of interest 1 vs. 2
pii6<-inv[3,3] #inv of cov, v1
pjj6<-inv[4,4] #inv of cov, v2
den1f<-seb3^2+seb4^2
den2f<-2*seb3*seb4
den3f<-pij6/(pii6+pjj6)
denc<-(den1f-(den2f*den3f))^.5
t34<-(b3-b4) / denc
p34<-2*(1-(pt(q=abs(t34), df=df, lower.tail=TRUE)))
t34<-round((t34),3)
message("Pred 1 vs. Pred 2 "," : t = ", t12,", p = ", p12)
message("Pred 1 vs. Pred 3 "," : t = ", t13,", p = ", p13)
message("Pred 1 vs. Pred 4 "," : t = ", t14,", p = ", p14)
message("Pred 2 vs. Pred 3 "," : t = ", t23,", p = ", p23)
message("Pred 2 vs. Pred 4 "," : t = ", t24,", p = ", p24)
message("Pred 3 vs. Pred 4 "," : t = ", t34,", p = ", p34)
}
if (numpred ==5 && comps=="raw")
{
xx<-dplyr::select(data,y,x1,x2,x3,x4,x5)
xx<-as.data.frame(xx)
mod<-lm(xx[,1]~xx[,2]+xx[,3]+xx[,4]+xx[,5]+xx[,6])
values<-summary(mod)
b1<-(values$coefficients)[2,1] #grabs b from each analysis
b2<-(values$coefficients)[3,1]
b3<-(values$coefficients)[4,1]
b4<-(values$coefficients)[5,1]
b5<-(values$coefficients)[6,1]
seb1<-(values$coefficients)[2,2]
seb2<-(values$coefficients)[3,2]
seb3<-(values$coefficients)[4,2]
seb4<-(values$coefficients)[5,2]
seb5<-(values$coefficients)[6,2]
df<- values$df[2]
r12<-cor(xx[,2],xx[,3])
r13<-cor(xx[,2],xx[,4])
r14<-cor(xx[,2],xx[,5])
r15<-cor(xx[,2],xx[,6])
r23<-cor(xx[,3],xx[,4])
r24<-cor(xx[,3],xx[,5])
r25<-cor(xx[,3],xx[,6])
r34<-cor(xx[,4],xx[,5])
r35<-cor(xx[,4],xx[,5])
r45<-cor(xx[,5],xx[,6])
mat<-cbind(c(1,r12,r13,r14,r15),c(r12,1,r23,r24,r25), c(r13,r23,1,r34,r35),
c(r14,r24,r34,1,r45),c(r15,r25,r35,r45,1))
inv<-solve(mat)*mat
# 1 vs 2
pij1<-inv[1,2] #inv of cor between pred of interest 1 vs. 2
pii1<-inv[1,1] #inv of cov, v1
pjj1<-inv[2,2] #inv of cov, v2
den1a<-seb1^2+seb2^2
den2a<-2*seb1*seb2
den3a<-pij1/(pii1+pjj1)
den<-(den1a-(den2a*den3a))^.5
t12<-(b1-b2) / den
p12<-2*(1-(pt(q=abs(t12), df=df, lower.tail=TRUE)))
t12<-round((t12),3)
#1 vs 3
pij2<-inv[1,3] #inv of cor between pred of interest 1 vs. 2
pii2<-inv[1,1] #inv of cov, v1
pjj2<-inv[3,3] #inv of cov, v2
den1b<-seb1^2+seb3^2
den2b<-2*seb1*seb3
den3b<-pij2/(pii2+pjj2)
denb<-(den1b-(den2b*den3b))^.5
t13<-(b1-b3) / denb
p13<-2*(1-(pt(q=abs(t13), df=df, lower.tail=TRUE)))
t13<-round((t13),3)
#1 vs 4
pij4<-inv[1,4] #inv of cor between pred of interest 1 vs. 2
pii4<-inv[1,1] #inv of cov, v1
pjj4<-inv[4,4] #inv of cov, v2
den1d<-seb1^2+seb4^2
den2d<-2*seb1*seb4
den3d<-pij4/(pii4+pjj4)
dend<-(den1d-(den2d*den3d))^.5
t14<-(b1-b4) / dend
p14<-2*(1-(pt(q=abs(t14), df=df, lower.tail=TRUE)))
t14<-round((t14),3)
#1 vs 5
pij6<-inv[1,5] #inv of cor between pred of interest 1 vs. 2
pii6<-inv[1,1] #inv of cov, v1
pjj6<-inv[5,5] #inv of cov, v2
den1f<-seb1^2+seb5^2
den2f<-2*seb1*seb5
den3f<-pij6/(pii4+pjj6)
denf<-(den1f-(den2f*den3f))^.5
t15<-(b1-b5) / dend
p15<-2*(1-(pt(q=abs(t15), df=df, lower.tail=TRUE)))
t15<-round((t15),3)
#2 vs 3
pij3<-inv[2,3] #inv of cor between pred of interest 1 vs. 2
pii3<-inv[2,2] #inv of cov, v1
pjj3<-inv[3,3] #inv of cov, v2
den1c<-seb2^2+seb3^2
den2c<-2*seb2*seb3
den3c<-pij3/(pii3+pjj3)
denc<-(den1c-(den2c*den3c))^.5
t23<-abs(abs(b2)-abs(b3)) / denc
p23<-2*(1-(pt(q=abs(t23), df=df, lower.tail=TRUE)))
t23<-round((t23),3)
#2 vs 4
pij5<-inv[2,4] #inv of cor between pred of interest 1 vs. 2
pii5<-inv[2,2] #inv of cov, v1
pjj5<-inv[4,4] #inv of cov, v2
den1e<-seb2^2+seb4^2
den2e<-2*seb2*seb4
den3e<-pij5/(pii5+pjj5)
denc<-(den1e-(den2e*den3e))^.5
t24<-(b2-b4) / denc
p24<-2*(1-(pt(q=abs(t24), df=df, lower.tail=TRUE)))
t24<-round((t24),3)
#2 vs 5
pij7<-inv[2,5] #inv of cor between pred of interest 1 vs. 2
pii7<-inv[2,2] #inv of cov, v1
pjj7<-inv[5,5] #inv of cov, v2
den1g<-seb2^2+seb5^2
den2g<-2*seb2*seb5
den3g<-pij7/(pii7+pjj7)
deng<-(den1g-(den2g*den3g))^.5
t25<-(b2-b5) / deng
p25<-2*(1-(pt(q=abs(t25), df=df, lower.tail=TRUE)))
t25<-round((t25),3)
#3 vs 4
pij8<-inv[3,4] #inv of cor between pred of interest 1 vs. 2
pii8<-inv[3,3] #inv of cov, v1
pjj8<-inv[4,4] #inv of cov, v2
den1h<-seb3^2+seb4^2
den2h<-2*seb3*seb4
den3h<-pij8/(pii8+pjj8)
denc<-(den1h-(den2h*den3h))^.5
t34<-(b3-b4) / denc
p34<-2*(1-(pt(q=abs(t34), df=df, lower.tail=TRUE)))
t34<-round((t34),3)
#3 vs 5
pij9<-inv[3,5] #inv of cor between pred of interest 1 vs. 2
pii9<-inv[3,3] #inv of cov, v1
pjj9<-inv[5,5] #inv of cov, v2
den1i<-seb3^2+seb5^2
den2i<-2*seb3*seb5
den3i<-pij9/(pii9+pjj9)
deni<-(den1i-(den2i*den3i))^.5
t35<-(b3-b5) / denc
p35<-2*(1-(pt(q=abs(t35), df=df, lower.tail=TRUE)))
t35<-round((t35),3)
#4 vs 5
pij10<-inv[4,5] #inv of cor between pred of interest 1 vs. 2
pii10<-inv[4,4] #inv of cov, v1
pjj10<-inv[5,5] #inv of cov, v2
den1j<-seb4^2+seb5^2
den2j<-2*seb4*seb5
den3j<-pij10/(pii10+pjj10)
denj<-(den1h-(den2h*den3h))^.5
t45<-(b4-b5) / denc
p45<-2*(1-(pt(q=abs(t45), df=df, lower.tail=TRUE)))
t45<-round((t45),3)
message("Pred 1 vs. Pred 2 "," : t = ", t12,", p = ", p12)
message("Pred 1 vs. Pred 3 "," : t = ", t13,", p = ", p13)
message("Pred 1 vs. Pred 4 "," : t = ", t14,", p = ", p14)
message("Pred 1 vs. Pred 5 "," : t = ", t15,", p = ", p15)
message("Pred 2 vs. Pred 3 "," : t = ", t23,", p = ", p23)
message("Pred 2 vs. Pred 4 "," : t = ", t24,", p = ", p24)
message("Pred 2 vs. Pred 5 "," : t = ", t25,", p = ", p25)
message("Pred 3 vs. Pred 4 "," : t = ", t34,", p = ", p34)
message("Pred 3 vs. Pred 5 "," : t = ", t35,", p = ", p35)
message("Pred 4 vs. Pred 5 "," : t = ", t45,", p = ", p45)
}
}
chrisaberson/BetterReg documentation built on June 10, 2022, 10:03 p.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.
Defines functions depbcomp
Documented in depbcomp
#'Power for Comparing Dependent Coefficients in Multiple Regression with Two or Three Predictors
#'Requires correlations between all variables as sample size. Means, sds, and alpha are option. Also computes Power(All)
#'@param data name of your datafile, loaded
#'@param y dependent variable name
#'@param x1 first predictor variable name
#'@param x2 second predictor variable name
#'@param x3 third predictor variable name
#'@param x4 fourth predictor variable name
#'@param x5 fifth predictor variable name
#'@param data name of data file
#'@param numpred number of predictors
#'@param comps Type of comparison, "abs" for absolute values or "raw" for raw coefficients
#'@examples
#'depbcomp(data=testreg,y="y",x1="x1",x2="x2",x3="x3",x4="x4",x5="x5", numpred=5,comps="abs")
#'@return Comparing Dependent Coefficients in Multiple Regression
#'@export
#'
depbcomp<-function(data=NULL,y=NULL, x1=NULL, x2=NULL,x3=NULL,x4=NULL,x5=NULL,numpred=NULL, comps="abs")
{
if (numpred ==2 && comps=="abs")
{
xx<-dplyr::select(data,y,x1,x2)
xx<-as.data.frame(xx)
mod<-lm(xx[,1]~xx[,2]+xx[,3])
values<-summary(mod)
b1<-(values$coefficients)[2,1] #grabs b from each analysis
b2<-(values$coefficients)[3,1]
seb1<-(values$coefficients)[2,2]
seb2<-(values$coefficients)[3,2]
df<- values$df[2]
varx1<-var(xx[,2])
varx2<-var(xx[,3])
r12<-abs(cor(xx[,2],xx[,3]))
mat<-cbind(c(1,r12),c(r12,1))
inv<-solve(mat)*mat
pij<-inv[1,2] #inv of cor between pred1 and 2
pii<-inv[1,1] #inv of cov, v1
pjj<-inv[2,2] #inv of cov, v2
den1<-seb1^2+seb2^2
den2<-2*seb1*seb2
den3<-pij/(pii+pjj)
den<-(den1-(den2*den3))^.5
t12<-abs((abs(b1)-(abs(b2)))) / den
p12<-2*(1-(pt(q=abs(t12), df=df, lower.tail=TRUE)))
t12<-round((t12),3)
message("Pred 1 vs. Pred 2 "," : t = ", t12,", p = ", p12)
}
if (numpred ==3 && comps=="abs")
{
xx<-dplyr::select(data,y,x1,x2,x3)
xx<-as.data.frame(xx)
mod<-lm(xx[,1]~xx[,2]+xx[,3]+xx[,4])
values<-summary(mod)
b1<-(values$coefficients)[2,1] #grabs b from each analysis
b2<-(values$coefficients)[3,1]
b3<-(values$coefficients)[4,1]
seb1<-(values$coefficients)[2,2]
seb2<-(values$coefficients)[3,2]
seb3<-(values$coefficients)[4,2]
df<- values$df[2]
r12<-abs(cor(xx[,2],xx[,3]))
r13<-abs(cor(xx[,2],xx[,4]))
r23<-abs(cor(xx[,3],xx[,4]))
mat<-cbind(c(1,r12,r13),c(r12,1,r23), c(r13,r23,1))
inv<-solve(mat)*mat
# 1 vs 2
pij1<-inv[1,2] #inv of cor between pred of interest 1 vs. 2
pii1<-inv[1,1] #inv of cov, v1
pjj1<-inv[2,2] #inv of cov, v2
den1a<-seb1^2+seb2^2
den2a<-2*seb1*seb2
den3a<-pij1/(pii1+pjj1)
den<-(den1a-(den2a*den3a))^.5
t12<-abs((abs(b1)-(abs(b2)))) / den
p12<-2*(1-(pt(q=abs(t12), df=df, lower.tail=TRUE)))
t12<-round((t12),3)
#1 vs 3
pij2<-inv[1,3] #inv of cor between pred of interest 1 vs. 2
pii2<-inv[1,1] #inv of cov, v1
pjj2<-inv[3,3] #inv of cov, v2
den1b<-seb1^2+seb3^2
den2b<-2*seb1*seb3
den3b<-pij2/(pii2+pjj2)
denb<-(den1b-(den2b*den3b))^.5
t13<-abs(abs(b1)-abs(b3)) / denb
p13<-2*(1-(pt(q=abs(t13), df=df, lower.tail=TRUE)))
t13<-round((t13),3)
#2 vs 3
pij3<-inv[2,3] #inv of cor between pred of interest 1 vs. 2
pii3<-inv[2,2] #inv of cov, v1
pjj3<-inv[3,3] #inv of cov, v2
den1c<-seb2^2+seb3^2
den2c<-2*seb2*seb3
den3c<-pij3/(pii3+pjj3)
denc<-(den1c-(den2c*den3c))^.5
t23<-abs(abs(b2)-abs(b3)) / denc
p23<-2*(1-(pt(q=abs(t23), df=df, lower.tail=TRUE)))
t23<-round((t23),3)
message("Pred 1 vs. Pred 2 "," : t = ", t12,", p = ", p12)
message("Pred 1 vs. Pred 3 "," : t = ", t13,", p = ", p13)
message("Pred 2 vs. Pred 3 "," : t = ", t23,", p = ", p23)
}
if (numpred ==4 && comps=="abs")
{
xx<-dplyr::select(data,y,x1,x2,x3,x4)
xx<-as.data.frame(xx)
mod<-lm(xx[,1]~xx[,2]+xx[,3]+xx[,4]+xx[,5])
values<-summary(mod)
b1<-(values$coefficients)[2,1] #grabs b from each analysis
b2<-(values$coefficients)[3,1]
b3<-(values$coefficients)[4,1]
b4<-(values$coefficients)[5,1]
seb1<-(values$coefficients)[2,2]
seb2<-(values$coefficients)[3,2]
seb3<-(values$coefficients)[4,2]
seb4<-(values$coefficients)[5,2]
df<- values$df[2]
r12<-abs(cor(xx[,2],xx[,3]))
r13<-abs(cor(xx[,2],xx[,4]))
r14<-abs(cor(xx[,2],xx[,5]))
r23<-abs(cor(xx[,3],xx[,4]))
r24<-abs(cor(xx[,3],xx[,5]))
r34<-abs(cor(xx[,4],xx[,5]))
mat<-cbind(c(1,r12,r13,r14),c(r12,1,r23,r24), c(r13,r23,1,r34),c(r14,r24,r34,1))
inv<-solve(mat)*mat
# 1 vs 2
pij1<-inv[1,2] #inv of cor between pred of interest 1 vs. 2
pii1<-inv[1,1] #inv of cov, v1
pjj1<-inv[2,2] #inv of cov, v2
den1a<-seb1^2+seb2^2
den2a<-2*seb1*seb2
den3a<-pij1/(pii1+pjj1)
den<-(den1a-(den2a*den3a))^.5
t12<-abs((abs(b1)-(abs(b2)))) / den
p12<-2*(1-(pt(q=abs(t12), df=df, lower.tail=TRUE)))
t12<-round((t12),3)
#1 vs 3
pij2<-inv[1,3] #inv of cor between pred of interest 1 vs. 2
pii2<-inv[1,1] #inv of cov, v1
pjj2<-inv[3,3] #inv of cov, v2
den1b<-seb1^2+seb3^2
den2b<-2*seb1*seb3
den3b<-pij2/(pii2+pjj2)
denb<-(den1b-(den2b*den3b))^.5
t13<-abs(abs(b1)-abs(b3)) / denb
p13<-2*(1-(pt(q=abs(t13), df=df, lower.tail=TRUE)))
t13<-round((t13),3)
#1 vs 4
pij4<-inv[1,4] #inv of cor between pred of interest 1 vs. 2
pii4<-inv[1,1] #inv of cov, v1
pjj4<-inv[4,4] #inv of cov, v2
den1d<-seb1^2+seb4^2
den2d<-2*seb1*seb4
den3d<-pij4/(pii4+pjj4)
dend<-(den1d-(den2d*den3d))^.5
t14<-abs(abs(b1)-abs(b4)) / dend
p14<-2*(1-(pt(q=abs(t14), df=df, lower.tail=TRUE)))
t14<-round((t14),3)
#2 vs 3
pij3<-inv[2,3] #inv of cor between pred of interest 1 vs. 2
pii3<-inv[2,2] #inv of cov, v1
pjj3<-inv[3,3] #inv of cov, v2
den1c<-seb2^2+seb3^2
den2c<-2*seb2*seb3
den3c<-pij3/(pii3+pjj3)
denc<-(den1c-(den2c*den3c))^.5
t23<-abs(abs(b2)-abs(b3)) / denc
p23<-2*(1-(pt(q=abs(t23), df=df, lower.tail=TRUE)))
t23<-round((t23),3)
#2 vs 4
pij5<-inv[2,4] #inv of cor between pred of interest 1 vs. 2
pii5<-inv[2,2] #inv of cov, v1
pjj5<-inv[4,4] #inv of cov, v2
den1e<-seb2^2+seb4^2
den2e<-2*seb2*seb4
den3e<-pij5/(pii5+pjj5)
denc<-(den1e-(den2e*den3e))^.5
t24<-abs(abs(b2)-abs(b4)) / denc
p24<-2*(1-(pt(q=abs(t24), df=df, lower.tail=TRUE)))
t24<-round((t24),3)
#3 vs 4
pij6<-inv[3,4] #inv of cor between pred of interest 1 vs. 2
pii6<-inv[3,3] #inv of cov, v1
pjj6<-inv[4,4] #inv of cov, v2
den1f<-seb3^2+seb4^2
den2f<-2*seb3*seb4
den3f<-pij6/(pii6+pjj6)
denc<-(den1f-(den2f*den3f))^.5
t34<-abs(abs(b3)-abs(b4)) / denc
p34<-2*(1-(pt(q=abs(t34), df=df, lower.tail=TRUE)))
t34<-round((t34),3)
message("Pred 1 vs. Pred 2 "," : t = ", t12,", p = ", p12)
message("Pred 1 vs. Pred 3 "," : t = ", t13,", p = ", p13)
message("Pred 1 vs. Pred 4 "," : t = ", t14,", p = ", p14)
message("Pred 2 vs. Pred 3 "," : t = ", t23,", p = ", p23)
message("Pred 2 vs. Pred 4 "," : t = ", t24,", p = ", p24)
message("Pred 3 vs. Pred 4 "," : t = ", t34,", p = ", p34)
}
if (numpred ==5 && comps=="abs")
{
xx<-dplyr::select(data,y,x1,x2,x3,x4,x5)
xx<-as.data.frame(xx)
mod<-lm(xx[,1]~xx[,2]+xx[,3]+xx[,4]+xx[,5]+xx[,6])
values<-summary(mod)
b1<-(values$coefficients)[2,1] #grabs b from each analysis
b2<-(values$coefficients)[3,1]
b3<-(values$coefficients)[4,1]
b4<-(values$coefficients)[5,1]
b5<-(values$coefficients)[6,1]
seb1<-(values$coefficients)[2,2]
seb2<-(values$coefficients)[3,2]
seb3<-(values$coefficients)[4,2]
seb4<-(values$coefficients)[5,2]
seb5<-(values$coefficients)[6,2]
df<- values$df[2]
r12<-abs(cor(xx[,2],xx[,3]))
r13<-abs(cor(xx[,2],xx[,4]))
r14<-abs(cor(xx[,2],xx[,5]))
r15<-abs(cor(xx[,2],xx[,6]))
r23<-abs(cor(xx[,3],xx[,4]))
r24<-abs(cor(xx[,3],xx[,5]))
r25<-abs(cor(xx[,3],xx[,6]))
r34<-abs(cor(xx[,4],xx[,5]))
r35<-abs(cor(xx[,4],xx[,5]))
r45<-abs(cor(xx[,5],xx[,6]))
mat<-cbind(c(1,r12,r13,r14,r15),c(r12,1,r23,r24,r25), c(r13,r23,1,r34,r35),
c(r14,r24,r34,1,r45),c(r15,r25,r35,r45,1))
inv<-solve(mat)*mat
# 1 vs 2
pij1<-inv[1,2] #inv of cor between pred of interest 1 vs. 2
pii1<-inv[1,1] #inv of cov, v1
pjj1<-inv[2,2] #inv of cov, v2
den1a<-seb1^2+seb2^2
den2a<-2*seb1*seb2
den3a<-pij1/(pii1+pjj1)
den<-(den1a-(den2a*den3a))^.5
t12<-abs((abs(b1)-(abs(b2)))) / den
p12<-2*(1-(pt(q=abs(t12), df=df, lower.tail=TRUE)))
t12<-round((t12),3)
#1 vs 3
pij2<-inv[1,3] #inv of cor between pred of interest 1 vs. 2
pii2<-inv[1,1] #inv of cov, v1
pjj2<-inv[3,3] #inv of cov, v2
den1b<-seb1^2+seb3^2
den2b<-2*seb1*seb3
den3b<-pij2/(pii2+pjj2)
denb<-(den1b-(den2b*den3b))^.5
t13<-abs(abs(b1)-abs(b3)) / denb
p13<-2*(1-(pt(q=abs(t13), df=df, lower.tail=TRUE)))
t13<-round((t13),3)
#1 vs 4
pij4<-inv[1,4] #inv of cor between pred of interest 1 vs. 2
pii4<-inv[1,1] #inv of cov, v1
pjj4<-inv[4,4] #inv of cov, v2
den1d<-seb1^2+seb4^2
den2d<-2*seb1*seb4
den3d<-pij4/(pii4+pjj4)
dend<-(den1d-(den2d*den3d))^.5
t14<-abs(abs(b1)-abs(b4)) / dend
p14<-2*(1-(pt(q=abs(t14), df=df, lower.tail=TRUE)))
t14<-round((t14),3)
#1 vs 5
pij6<-inv[1,5] #inv of cor between pred of interest 1 vs. 2
pii6<-inv[1,1] #inv of cov, v1
pjj6<-inv[5,5] #inv of cov, v2
den1f<-seb1^2+seb5^2
den2f<-2*seb1*seb5
den3f<-pij6/(pii4+pjj6)
denf<-(den1f-(den2f*den3f))^.5
t15<-abs(abs(b1)-abs(b5)) / dend
p15<-2*(1-(pt(q=abs(t15), df=df, lower.tail=TRUE)))
t15<-round((t15),3)
#2 vs 3
pij3<-inv[2,3] #inv of cor between pred of interest 1 vs. 2
pii3<-inv[2,2] #inv of cov, v1
pjj3<-inv[3,3] #inv of cov, v2
den1c<-seb2^2+seb3^2
den2c<-2*seb2*seb3
den3c<-pij3/(pii3+pjj3)
denc<-(den1c-(den2c*den3c))^.5
t23<-abs(abs(b2)-abs(b3)) / denc
p23<-2*(1-(pt(q=abs(t23), df=df, lower.tail=TRUE)))
t23<-round((t23),3)
#2 vs 4
pij5<-inv[2,4] #inv of cor between pred of interest 1 vs. 2
pii5<-inv[2,2] #inv of cov, v1
pjj5<-inv[4,4] #inv of cov, v2
den1e<-seb2^2+seb4^2
den2e<-2*seb2*seb4
den3e<-pij5/(pii5+pjj5)
denc<-(den1e-(den2e*den3e))^.5
t24<-abs(abs(b2)-abs(b4)) / denc
p24<-2*(1-(pt(q=abs(t24), df=df, lower.tail=TRUE)))
t24<-round((t24),3)
#2 vs 5
pij7<-inv[2,5] #inv of cor between pred of interest 1 vs. 2
pii7<-inv[2,2] #inv of cov, v1
pjj7<-inv[5,5] #inv of cov, v2
den1g<-seb2^2+seb5^2
den2g<-2*seb2*seb5
den3g<-pij7/(pii7+pjj7)
deng<-(den1g-(den2g*den3g))^.5
t25<-abs(abs(b2)-abs(b5)) / deng
p25<-2*(1-(pt(q=abs(t25), df=df, lower.tail=TRUE)))
t25<-round((t25),3)
#3 vs 4
pij8<-inv[3,4] #inv of cor between pred of interest 1 vs. 2
pii8<-inv[3,3] #inv of cov, v1
pjj8<-inv[4,4] #inv of cov, v2
den1h<-seb3^2+seb4^2
den2h<-2*seb3*seb4
den3h<-pij8/(pii8+pjj8)
denc<-(den1h-(den2h*den3h))^.5
t34<-abs(abs(b3)-abs(b4)) / denc
p34<-2*(1-(pt(q=abs(t34), df=df, lower.tail=TRUE)))
t34<-round((t34),3)
#3 vs 5
pij9<-inv[3,5] #inv of cor between pred of interest 1 vs. 2
pii9<-inv[3,3] #inv of cov, v1
pjj9<-inv[5,5] #inv of cov, v2
den1i<-seb3^2+seb5^2
den2i<-2*seb3*seb5
den3i<-pij9/(pii9+pjj9)
deni<-(den1i-(den2i*den3i))^.5
t35<-abs(abs(b3)-abs(b5)) / denc
p35<-2*(1-(pt(q=abs(t35), df=df, lower.tail=TRUE)))
t35<-round((t35),3)
#4 vs 5
pij10<-inv[4,5] #inv of cor between pred of interest 1 vs. 2
pii10<-inv[4,4] #inv of cov, v1
pjj10<-inv[5,5] #inv of cov, v2
den1j<-seb4^2+seb5^2
den2j<-2*seb4*seb5
den3j<-pij10/(pii10+pjj10)
denj<-(den1h-(den2h*den3h))^.5
t45<-abs(abs(b4)-abs(b5)) / denc
p45<-2*(1-(pt(q=abs(t45), df=df, lower.tail=TRUE)))
t45<-round((t45),3)
message("Pred 1 vs. Pred 2 "," : t = ", t12,", p = ", p12)
message("Pred 1 vs. Pred 3 "," : t = ", t13,", p = ", p13)
message("Pred 1 vs. Pred 4 "," : t = ", t14,", p = ", p14)
message("Pred 1 vs. Pred 5 "," : t = ", t15,", p = ", p15)
message("Pred 2 vs. Pred 3 "," : t = ", t23,", p = ", p23)
message("Pred 2 vs. Pred 4 "," : t = ", t24,", p = ", p24)
message("Pred 2 vs. Pred 5 "," : t = ", t25,", p = ", p25)
message("Pred 3 vs. Pred 4 "," : t = ", t34,", p = ", p34)
message("Pred 3 vs. Pred 5 "," : t = ", t35,", p = ", p35)
message("Pred 4 vs. Pred 5 "," : t = ", t45,", p = ", p45)
}
if (numpred ==2 && comps=="raw")
{
xx<-dplyr::select(data,y,x1,x2)
xx<-as.data.frame(xx)
mod<-lm(xx[,1]~xx[,2]+xx[,3])
values<-summary(mod)
b1<-(values$coefficients)[2,1] #grabs b from each analysis
b2<-(values$coefficients)[3,1]
seb1<-(values$coefficients)[2,2]
seb2<-(values$coefficients)[3,2]
df<- values$df[2]
varx1<-var(xx[,2])
varx2<-var(xx[,3])
r12<-cor(xx[,2],xx[,3])
mat<-cbind(c(1,r12),c(r12,1))
inv<-solve(mat)*mat
pij<-inv[1,2] #inv of cor between pred1 and 2
pii<-inv[1,1] #inv of cov, v1
pjj<-inv[2,2] #inv of cov, v2
den1<-seb1^2+seb2^2
den2<-2*seb1*seb2
den3<-pij/(pii+pjj)
den<-(den1-(den2*den3))^.5
t12<-(b1-b2)/ den
p12<-2*(1-(pt(q=abs(t12), df=df, lower.tail=TRUE)))
t12<-round((t12),3)
message("Pred 1 vs. Pred 2 "," : t = ", t12,", p = ", p12)
}
if (numpred ==3 && comps=="raw")
{
xx<-dplyr::select(data,y,x1,x2,x3)
xx<-as.data.frame(xx)
mod<-lm(xx[,1]~xx[,2]+xx[,3]+xx[,4])
values<-summary(mod)
b1<-(values$coefficients)[2,1] #grabs b from each analysis
b2<-(values$coefficients)[3,1]
b3<-(values$coefficients)[4,1]
seb1<-(values$coefficients)[2,2]
seb2<-(values$coefficients)[3,2]
seb3<-(values$coefficients)[4,2]
df<- values$df[2]
r12<-cor(xx[,2],xx[,3])
r13<-cor(xx[,2],xx[,4])
r23<-cor(xx[,3],xx[,4])
mat<-cbind(c(1,r12,r13),c(r12,1,r23), c(r13,r23,1))
inv<-solve(mat)*mat
# 1 vs 2
pij1<-inv[1,2] #inv of cor between pred of interest 1 vs. 2
pii1<-inv[1,1] #inv of cov, v1
pjj1<-inv[2,2] #inv of cov, v2
den1a<-seb1^2+seb2^2
den2a<-2*seb1*seb2
den3a<-pij1/(pii1+pjj1)
den<-(den1a-(den2a*den3a))^.5
t12<-(b1-b2) / den
p12<-2*(1-(pt(q=abs(t12), df=df, lower.tail=TRUE)))
t12<-round((t12),3)
#1 vs 3
pij2<-inv[1,3] #inv of cor between pred of interest 1 vs. 2
pii2<-inv[1,1] #inv of cov, v1
pjj2<-inv[3,3] #inv of cov, v2
den1b<-seb1^2+seb3^2
den2b<-2*seb1*seb3
den3b<-pij2/(pii2+pjj2)
denb<-(den1b-(den2b*den3b))^.5
t13<-(b1-b3) / denb
p13<-2*(1-(pt(q=abs(t13), df=df, lower.tail=TRUE)))
t13<-round((t13),3)
#2 vs 3
pij3<-inv[2,3] #inv of cor between pred of interest 1 vs. 2
pii3<-inv[2,2] #inv of cov, v1
pjj3<-inv[3,3] #inv of cov, v2
den1c<-seb2^2+seb3^2
den2c<-2*seb2*seb3
den3c<-pij3/(pii3+pjj3)
denc<-(den1c-(den2c*den3c))^.5
t23<-abs(abs(b2)-abs(b3)) / denc
p23<-2*(1-(pt(q=abs(t23), df=df, lower.tail=TRUE)))
t23<-round((t23),3)
message("Pred 1 vs. Pred 2 "," : t = ", t12,", p = ", p12)
message("Pred 1 vs. Pred 3 "," : t = ", t13,", p = ", p13)
message("Pred 2 vs. Pred 3 "," : t = ", t23,", p = ", p23)
}
if (numpred ==4 && comps=="raw")
{
xx<-dplyr::select(data,y,x1,x2,x3,x4)
xx<-as.data.frame(xx)
mod<-lm(xx[,1]~xx[,2]+xx[,3]+xx[,4]+xx[,5])
values<-summary(mod)
b1<-(values$coefficients)[2,1] #grabs b from each analysis
b2<-(values$coefficients)[3,1]
b3<-(values$coefficients)[4,1]
b4<-(values$coefficients)[5,1]
seb1<-(values$coefficients)[2,2]
seb2<-(values$coefficients)[3,2]
seb3<-(values$coefficients)[4,2]
seb4<-(values$coefficients)[5,2]
df<- values$df[2]
r12<-cor(xx[,2],xx[,3])
r13<-cor(xx[,2],xx[,4])
r14<-cor(xx[,2],xx[,5])
r23<-cor(xx[,3],xx[,4])
r24<-cor(xx[,3],xx[,5])
r34<-cor(xx[,4],xx[,5])
mat<-cbind(c(1,r12,r13,r14),c(r12,1,r23,r24), c(r13,r23,1,r34),c(r14,r24,r34,1))
inv<-solve(mat)*mat
# 1 vs 2
pij1<-inv[1,2] #inv of cor between pred of interest 1 vs. 2
pii1<-inv[1,1] #inv of cov, v1
pjj1<-inv[2,2] #inv of cov, v2
den1a<-seb1^2+seb2^2
den2a<-2*seb1*seb2
den3a<-pij1/(pii1+pjj1)
den<-(den1a-(den2a*den3a))^.5
t12<-(b1-b2) / den
p12<-2*(1-(pt(q=abs(t12), df=df, lower.tail=TRUE)))
t12<-round((t12),3)
#1 vs 3
pij2<-inv[1,3] #inv of cor between pred of interest 1 vs. 2
pii2<-inv[1,1] #inv of cov, v1
pjj2<-inv[3,3] #inv of cov, v2
den1b<-seb1^2+seb3^2
den2b<-2*seb1*seb3
den3b<-pij2/(pii2+pjj2)
denb<-(den1b-(den2b*den3b))^.5
t13<-(b1-b3) / denb
p13<-2*(1-(pt(q=abs(t13), df=df, lower.tail=TRUE)))
t13<-round((t13),3)
#1 vs 4
pij4<-inv[1,4] #inv of cor between pred of interest 1 vs. 2
pii4<-inv[1,1] #inv of cov, v1
pjj4<-inv[4,4] #inv of cov, v2
den1d<-seb1^2+seb4^2
den2d<-2*seb1*seb4
den3d<-pij4/(pii4+pjj4)
dend<-(den1d-(den2d*den3d))^.5
t14<-(b1-b4) / dend
p14<-2*(1-(pt(q=abs(t14), df=df, lower.tail=TRUE)))
t14<-round((t14),3)
#2 vs 3
pij3<-inv[2,3] #inv of cor between pred of interest 1 vs. 2
pii3<-inv[2,2] #inv of cov, v1
pjj3<-inv[3,3] #inv of cov, v2
den1c<-seb2^2+seb3^2
den2c<-2*seb2*seb3
den3c<-pij3/(pii3+pjj3)
denc<-(den1c-(den2c*den3c))^.5
t23<-abs(abs(b2)-abs(b3)) / denc
p23<-2*(1-(pt(q=abs(t23), df=df, lower.tail=TRUE)))
t23<-round((t23),3)
#2 vs 4
pij5<-inv[2,4] #inv of cor between pred of interest 1 vs. 2
pii5<-inv[2,2] #inv of cov, v1
pjj5<-inv[4,4] #inv of cov, v2
den1e<-seb2^2+seb4^2
den2e<-2*seb2*seb4
den3e<-pij5/(pii5+pjj5)
denc<-(den1e-(den2e*den3e))^.5
t24<-(b2-b4) / denc
p24<-2*(1-(pt(q=abs(t24), df=df, lower.tail=TRUE)))
t24<-round((t24),3)
#3 vs 4
pij6<-inv[3,4] #inv of cor between pred of interest 1 vs. 2
pii6<-inv[3,3] #inv of cov, v1
pjj6<-inv[4,4] #inv of cov, v2
den1f<-seb3^2+seb4^2
den2f<-2*seb3*seb4
den3f<-pij6/(pii6+pjj6)
denc<-(den1f-(den2f*den3f))^.5
t34<-(b3-b4) / denc
p34<-2*(1-(pt(q=abs(t34), df=df, lower.tail=TRUE)))
t34<-round((t34),3)
message("Pred 1 vs. Pred 2 "," : t = ", t12,", p = ", p12)
message("Pred 1 vs. Pred 3 "," : t = ", t13,", p = ", p13)
message("Pred 1 vs. Pred 4 "," : t = ", t14,", p = ", p14)
message("Pred 2 vs. Pred 3 "," : t = ", t23,", p = ", p23)
message("Pred 2 vs. Pred 4 "," : t = ", t24,", p = ", p24)
message("Pred 3 vs. Pred 4 "," : t = ", t34,", p = ", p34)
}
if (numpred ==5 && comps=="raw")
{
xx<-dplyr::select(data,y,x1,x2,x3,x4,x5)
xx<-as.data.frame(xx)
mod<-lm(xx[,1]~xx[,2]+xx[,3]+xx[,4]+xx[,5]+xx[,6])
values<-summary(mod)
b1<-(values$coefficients)[2,1] #grabs b from each analysis
b2<-(values$coefficients)[3,1]
b3<-(values$coefficients)[4,1]
b4<-(values$coefficients)[5,1]
b5<-(values$coefficients)[6,1]
seb1<-(values$coefficients)[2,2]
seb2<-(values$coefficients)[3,2]
seb3<-(values$coefficients)[4,2]
seb4<-(values$coefficients)[5,2]
seb5<-(values$coefficients)[6,2]
df<- values$df[2]
r12<-cor(xx[,2],xx[,3])
r13<-cor(xx[,2],xx[,4])
r14<-cor(xx[,2],xx[,5])
r15<-cor(xx[,2],xx[,6])
r23<-cor(xx[,3],xx[,4])
r24<-cor(xx[,3],xx[,5])
r25<-cor(xx[,3],xx[,6])
r34<-cor(xx[,4],xx[,5])
r35<-cor(xx[,4],xx[,5])
r45<-cor(xx[,5],xx[,6])
mat<-cbind(c(1,r12,r13,r14,r15),c(r12,1,r23,r24,r25), c(r13,r23,1,r34,r35),
c(r14,r24,r34,1,r45),c(r15,r25,r35,r45,1))
inv<-solve(mat)*mat
# 1 vs 2
pij1<-inv[1,2] #inv of cor between pred of interest 1 vs. 2
pii1<-inv[1,1] #inv of cov, v1
pjj1<-inv[2,2] #inv of cov, v2
den1a<-seb1^2+seb2^2
den2a<-2*seb1*seb2
den3a<-pij1/(pii1+pjj1)
den<-(den1a-(den2a*den3a))^.5
t12<-(b1-b2) / den
p12<-2*(1-(pt(q=abs(t12), df=df, lower.tail=TRUE)))
t12<-round((t12),3)
#1 vs 3
pij2<-inv[1,3] #inv of cor between pred of interest 1 vs. 2
pii2<-inv[1,1] #inv of cov, v1
pjj2<-inv[3,3] #inv of cov, v2
den1b<-seb1^2+seb3^2
den2b<-2*seb1*seb3
den3b<-pij2/(pii2+pjj2)
denb<-(den1b-(den2b*den3b))^.5
t13<-(b1-b3) / denb
p13<-2*(1-(pt(q=abs(t13), df=df, lower.tail=TRUE)))
t13<-round((t13),3)
#1 vs 4
pij4<-inv[1,4] #inv of cor between pred of interest 1 vs. 2
pii4<-inv[1,1] #inv of cov, v1
pjj4<-inv[4,4] #inv of cov, v2
den1d<-seb1^2+seb4^2
den2d<-2*seb1*seb4
den3d<-pij4/(pii4+pjj4)
dend<-(den1d-(den2d*den3d))^.5
t14<-(b1-b4) / dend
p14<-2*(1-(pt(q=abs(t14), df=df, lower.tail=TRUE)))
t14<-round((t14),3)
#1 vs 5
pij6<-inv[1,5] #inv of cor between pred of interest 1 vs. 2
pii6<-inv[1,1] #inv of cov, v1
pjj6<-inv[5,5] #inv of cov, v2
den1f<-seb1^2+seb5^2
den2f<-2*seb1*seb5
den3f<-pij6/(pii4+pjj6)
denf<-(den1f-(den2f*den3f))^.5
t15<-(b1-b5) / dend
p15<-2*(1-(pt(q=abs(t15), df=df, lower.tail=TRUE)))
t15<-round((t15),3)
#2 vs 3
pij3<-inv[2,3] #inv of cor between pred of interest 1 vs. 2
pii3<-inv[2,2] #inv of cov, v1
pjj3<-inv[3,3] #inv of cov, v2
den1c<-seb2^2+seb3^2
den2c<-2*seb2*seb3
den3c<-pij3/(pii3+pjj3)
denc<-(den1c-(den2c*den3c))^.5
t23<-abs(abs(b2)-abs(b3)) / denc
p23<-2*(1-(pt(q=abs(t23), df=df, lower.tail=TRUE)))
t23<-round((t23),3)
#2 vs 4
pij5<-inv[2,4] #inv of cor between pred of interest 1 vs. 2
pii5<-inv[2,2] #inv of cov, v1
pjj5<-inv[4,4] #inv of cov, v2
den1e<-seb2^2+seb4^2
den2e<-2*seb2*seb4
den3e<-pij5/(pii5+pjj5)
denc<-(den1e-(den2e*den3e))^.5
t24<-(b2-b4) / denc
p24<-2*(1-(pt(q=abs(t24), df=df, lower.tail=TRUE)))
t24<-round((t24),3)
#2 vs 5
pij7<-inv[2,5] #inv of cor between pred of interest 1 vs. 2
pii7<-inv[2,2] #inv of cov, v1
pjj7<-inv[5,5] #inv of cov, v2
den1g<-seb2^2+seb5^2
den2g<-2*seb2*seb5
den3g<-pij7/(pii7+pjj7)
deng<-(den1g-(den2g*den3g))^.5
t25<-(b2-b5) / deng
p25<-2*(1-(pt(q=abs(t25), df=df, lower.tail=TRUE)))
t25<-round((t25),3)
#3 vs 4
pij8<-inv[3,4] #inv of cor between pred of interest 1 vs. 2
pii8<-inv[3,3] #inv of cov, v1
pjj8<-inv[4,4] #inv of cov, v2
den1h<-seb3^2+seb4^2
den2h<-2*seb3*seb4
den3h<-pij8/(pii8+pjj8)
denc<-(den1h-(den2h*den3h))^.5
t34<-(b3-b4) / denc
p34<-2*(1-(pt(q=abs(t34), df=df, lower.tail=TRUE)))
t34<-round((t34),3)
#3 vs 5
pij9<-inv[3,5] #inv of cor between pred of interest 1 vs. 2
pii9<-inv[3,3] #inv of cov, v1
pjj9<-inv[5,5] #inv of cov, v2
den1i<-seb3^2+seb5^2
den2i<-2*seb3*seb5
den3i<-pij9/(pii9+pjj9)
deni<-(den1i-(den2i*den3i))^.5
t35<-(b3-b5) / denc
p35<-2*(1-(pt(q=abs(t35), df=df, lower.tail=TRUE)))
t35<-round((t35),3)
#4 vs 5
pij10<-inv[4,5] #inv of cor between pred of interest 1 vs. 2
pii10<-inv[4,4] #inv of cov, v1
pjj10<-inv[5,5] #inv of cov, v2
den1j<-seb4^2+seb5^2
den2j<-2*seb4*seb5
den3j<-pij10/(pii10+pjj10)
denj<-(den1h-(den2h*den3h))^.5
t45<-(b4-b5) / denc
p45<-2*(1-(pt(q=abs(t45), df=df, lower.tail=TRUE)))
t45<-round((t45),3)
message("Pred 1 vs. Pred 2 "," : t = ", t12,", p = ", p12)
message("Pred 1 vs. Pred 3 "," : t = ", t13,", p = ", p13)
message("Pred 1 vs. Pred 4 "," : t = ", t14,", p = ", p14)
message("Pred 1 vs. Pred 5 "," : t = ", t15,", p = ", p15)
message("Pred 2 vs. Pred 3 "," : t = ", t23,", p = ", p23)
message("Pred 2 vs. Pred 4 "," : t = ", t24,", p = ", p24)
message("Pred 2 vs. Pred 5 "," : t = ", t25,", p = ", p25)
message("Pred 3 vs. Pred 4 "," : t = ", t34,", p = ", p34)
message("Pred 3 vs. Pred 5 "," : t = ", t35,", p = ", p35)
message("Pred 4 vs. Pred 5 "," : t = ", t45,", p = ", p45)
}
}
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.