How to loop multiple exposures and outcomes with glm in R?

Question

I have a loop that currently works to test multiple exposures with one outcome in R.

The code below tests associations for outcome y with exp1, exp2, and exp3.

My question is, what would be the best/efficient way to test the same exposure associations for y, y1, y2, y3, y4? I am trying to run glm for multiple exposures and multiple outcomes. Instead of me copying out the loop 5 times for the 5 outcomes.

# Build data --------------------------------------------------------------
amino_df <- data.frame(y = rbinom(100, 1, 0.5), y2 = rbinom(100, 1, 0.3), y3 = rbinom(100, 1, 0.2), y4 = rbinom(100, 1, 0.22),
                       exp1 = rnorm(100), exp2 = rnorm(100), exp3 = rnorm(100))

# Observational estimates unadjusted -------------------------------------------------

exp <- c("exp1", "exp2", "exp3")

obs_results <- data.frame()  

for (i in seq_along(exp))
{
  mod <- as.formula(sprintf("y ~ %s", exp[i]))
  glmmodel <- glm(formula = mod, family = binomial, data = amino_df)
  
  obs_results[i,1] <- names(coef(glmmodel))[2]
  obs_results[i,2] <- exp(glmmodel$coefficients[2])
  obs_results[i,3] <- summary(glmmodel)$coefficients[2,2]
  obs_results[i,4] <- summary(glmmodel)$coefficients[2,4]
  obs_results[i,5] <- exp(confint.default(glmmodel)[2,1])
  obs_results[i,6] <- exp(confint.default(glmmodel)[2,2])
  
  colnames(obs_results) <- c("exposure","OR", "SE", "P_value", "95_CI_LOW","95_CI_HIGH")
}

Answer 1

The same thing that Elena did, but using lists:

exp <- c("exp1", "exp2", "exp3")
y <- c("y","y2","y3")

obs_results <- replicate(length(y), data.frame())  

for(j in seq_along(y)){
  for (i in seq_along(exp)){
    mod <- as.formula(paste(y[j], "~", exp[i]))
    glmmodel <- glm(formula = mod, family = binomial, data = amino_df)
    
    obs_results[[j]][i,1] <- names(coef(glmmodel))[2]
    obs_results[[j]][i,2] <- exp(glmmodel$coefficients[2])
    obs_results[[j]][i,3] <- summary(glmmodel)$coefficients[2,2]
    obs_results[[j]][i,4] <- summary(glmmodel)$coefficients[2,4]
    obs_results[[j]][i,5] <- exp(confint.default(glmmodel)[2,1])
    obs_results[[j]][i,6] <- exp(confint.default(glmmodel)[2,2])
  }
  colnames(obs_results[[j]]) <- c("exposure","OR", "SE", "P_value", "95_CI_LOW","95_CI_HIGH")
}
names(obs_results) <- y

Output:

> obs_results
$y
  exposure       OR        SE   P_value 95_CI_LOW 95_CI_HIGH
1     exp1 0.992145 0.2023656 0.9689149 0.6673001   1.475126
2     exp2 1.064498 0.2107148 0.7667543 0.7043425   1.608812
3     exp3 0.704014 0.2143235 0.1015239 0.4625395   1.071553

$y2
  exposure        OR        SE   P_value 95_CI_LOW 95_CI_HIGH
1     exp1 0.9246032 0.2260353 0.7287363 0.5936818   1.439982
2     exp2 0.8905785 0.2347429 0.6215439 0.5621584   1.410866
3     exp3 1.2104091 0.2299170 0.4062258 0.7713056   1.899494

$y3
  exposure        OR        SE   P_value 95_CI_LOW 95_CI_HIGH
1     exp1 1.1224366 0.2425520 0.6339361 0.6977522   1.805604
2     exp2 0.9870573 0.2532694 0.9589780 0.6008403   1.621533
3     exp3 0.6854464 0.2582983 0.1436851 0.4131517   1.137201

Answer 2

You can simply wrap another loop around it:

exp <- c("exp1", "exp2", "exp3")
ys <- c("y2","y3","y4")

obs_results_total <- data.frame() 
obs_results <- data.frame()  
for (j in ys){

for (i in seq_along(exp))
{
  mod <- as.formula(sprintf("%s ~ %s",j ,exp[i]))
  glmmodel <- glm(formula = mod, family = binomial, data = amino_df)
  
  obs_results[i,1] <- names(coef(glmmodel))[2]
  obs_results[i,2] <- exp(glmmodel$coefficients[2])
  obs_results[i,3] <- summary(glmmodel)$coefficients[2,2]
  obs_results[i,4] <- summary(glmmodel)$coefficients[2,4]
  obs_results[i,5] <- exp(confint.default(glmmodel)[2,1])
  obs_results[i,6] <- exp(confint.default(glmmodel)[2,2])
  obs_results[i,7] <- j
  
  colnames(obs_results) <- c("exposure","OR", "SE", "P_value", "95_CI_LOW","95_CI_HIGH","y")
}
  obs_results_total <- rbind(obs_results_total,obs_results)
}