making function more universal

R/rep_glm.R

@@ -1,33 +1,39 @@
 #' A repeated logistic regression function
 #'
-#' For bivariate analyses.
+#' @description For bivariate analyses. The confint() function is rather slow, causing the whole function to hang when including many predictors and calculating the ORs with CI.
 #' @param y Effect meassure.
-#' @param v1 Main variable in model
+#' @param vars variables in model. Input as c() of columnnames, use dput().
+#' @param string variables to test. Input as c() of columnnames, use dput().
 #' @keywords logistic regression
 #' @export
 #' @examples
 #' rep_glm()
 
-rep_glm<-function(y,v1,string,ci=FALSE,data,v2=NULL,v3=NULL){
+rep_glm<-function(y,vars,string,ci=FALSE,data){
   ## x is data.frame of predictors, y is vector of an aoutcome as a factor
   ## output is returned as coefficient, or if or=TRUE as OR with 95 % CI.
-## The confint() function is rather slow, causing the whole function to hang when including many predictors and calculating the ORs with CI.
+  ##
 
-  require(broom)
+  require(dplyr)
 
   d<-data
   x<-select(d,one_of(c(string)))
-  m1<-length(coef(glm(y~v1,family = binomial())))
+  v<-select(d,one_of(c(vars)))
+  dt<-cbind(y,v)
+  m1<-length(coef(glm(y~.,family = binomial(),data = dt)))
 
   if (!is.factor(y)){stop("Some kind of error message would be nice, but y should be a factor!")}
 
   if (ci==TRUE){
 
-             df<-data.frame(matrix(ncol = 4))
-  names(df)<-c("pred","or_ci","pv","t")
+    df<-data.frame(matrix(ncol = 3))
+    names(df)<-c("pred","or_ci","pv")
 
     for(i in 1:ncol(x)){
-     m<-glm(y~v1+x[,i],family = binomial())
+
+      dat<-cbind(dt,x[,i])
+
+      m<-glm(y~.,family = binomial(),data=dat)
 
       l<-suppressMessages(round(exp(confint(m))[-c(1:m1),1],2))
       u<-suppressMessages(round(exp(confint(m))[-c(1:m1),2],2))
@@ -36,56 +42,54 @@ if (!is.factor(y)){stop("Some kind of error message would be nice, but y should
       or_ci<-paste0(or," (",l," to ",u,")")
 
       pv<-round(tidy(m)$p.value[-c(1:m1)],3)
-     pv<-ifelse(pv<0.001,"<0.001",pv)
 
-         t <- ifelse(pv<=0.1|pv=="<0.001","include","drop")
+      x1<-x[,i]
 
-     pv <- ifelse(pv<=0.05|pv=="<0.001",paste0("*",pv),
-                  ifelse(pv>0.05&pv<=0.1,paste0(".",pv),pv))
-
-
-    v<-x[,i]
-
-    if (is.factor(v)){
-       pred<-paste(names(x)[i],levels(v)[-1],sep = "_")
+      if (is.factor(x1)){
+        pred<-paste(names(x)[i],levels(x1)[-1],sep = "_")
       }
 
       else {pred<-names(x)[i]}
 
-      df<-rbind(df,cbind(pred,or_ci,pv,t))
+      df<-rbind(df,cbind(pred,or_ci,pv))
 
     }}
 
   if (ci==FALSE){
 
-       df<-data.frame(matrix(ncol = 4))
-  names(df)<-c("pred","b","pv","t")
+    df<-data.frame(matrix(ncol = 3))
+    names(df)<-c("pred","b","pv")
 
     for(i in 1:ncol(x)){
-     m<-glm(y~v1+x[,i],family = binomial())
+      dat<-cbind(dt,x[,i])
+
+      m<-glm(y~.,family = binomial(),data=dat)
 
       b<-round(coef(m)[-c(1:m1)],3)
 
       pv<-round(tidy(m)$p.value[-c(1:m1)],3)
-     pv<-ifelse(pv<0.001,"<0.001",pv)
 
-    t <- ifelse(pv<=0.1|pv=="<0.001","include","drop")
+      x1<-x[,i]
 
-     pv <- ifelse(pv<=0.05|pv=="<0.001",paste0("*",pv),
-                  ifelse(pv>0.05&pv<=0.1,paste0(".",pv),pv))
-
-
-     v<-x[,i]
-
-     if (is.factor(v)){
-       pred<-paste(names(x)[i],levels(v)[-1],sep = "_")
+      if (is.factor(x1)){
+        pred<-paste(names(x1)[i],levels(x1)[-1],sep = "_")
       }
 
       else {pred<-names(x)[i]}
 
-     df<-rbind(df,cbind(pred,b,pv,t))
+      df<-rbind(df,cbind(pred,b,pv))
 
     }}
-  result<-df
-  return(df)
+
+  pa<-as.numeric(df[,3])
+  pa<-ifelse(pa<0.001,"<0.001",pa)
+
+  t <- ifelse(pa<=0.1|pa=="<0.001","include","drop")
+
+  pa <- ifelse(pa<=0.05|pa=="<0.001",paste0("*",pa),
+               ifelse(pa>0.05&pa<=0.1,paste0(".",pa),pa))
+
+  r<-data.frame(df[,1:2],pa,t)[-1,]
+
+  return(r)
 }

man/rep_glm.Rd

@@ -4,15 +4,17 @@
 \alias{rep_glm}
 \title{A repeated logistic regression function}
 \usage{
-rep_glm(y, v1, string, ci = FALSE, data, v2 = NULL, v3 = NULL)
+rep_glm(y, vars, string, ci = FALSE, data)
 }
 \arguments{
 \item{y}{Effect meassure.}
 
-\item{v1}{Main variable in model}
+\item{vars}{variables in model. Input as c() of columnnames, use dput().}
+
+\item{string}{variables to test. Input as c() of columnnames, use dput().}
 }
 \description{
-For bivariate analyses.
+For bivariate analyses. The confint() function is rather slow, causing the whole function to hang when including many predictors and calculating the ORs with CI.
 }
 \examples{
 rep_glm()