How to create a function that calculates multiple statistical functions?-CodePudding

I want to make statistical analysis of many vectors, such as cor, MAE, bias, sd, t.test, chisq.test,... and I guess there is any way of creating a function that I only feed the data to analyze and it gives a vector with the calculations.

Ideally I would like to feed vector1 and vector2, and get the calculations made.

Right now I am doint the following, but it is gettin unsustainable pretty fast.

  ## R^2
rsq_15_18 <- round(cor(x = study_15_18$potential_15, y = study_15_18$overall_18 ,method = "pearson")^2,4)
rsq_16_19 <- round(cor(x = study_16_19$potential_16, y = study_16_19$overall_19 ,method = "pearson")^2,4)
rsq_17_20 <- round(cor(x = study_17_20$potential_17, y = study_17_20$overall_20 ,method = "pearson")^2,4)
rsq_18_21 <- round(cor(x = study_18_21$potential_18, y = study_18_21$overall_21 ,method = "pearson")^2,4)
rsq_19_22 <- round(cor(x = study_19_22$potential_19, y = study_19_22$overall_22 ,method = "pearson")^2,4)

## MAE
mae_15_18 <- round(mae(study_15_18$overall_18, study_15_18$potential_15),4)
mae_16_19 <- round(mae(study_16_19$overall_19, study_16_19$potential_16),4)
mae_17_20 <- round(mae(study_17_20$overall_20, study_17_20$potential_17),4)
mae_18_21 <- round(mae(study_18_21$overall_21, study_18_21$potential_18),4)
mae_19_22 <- round(mae(study_19_22$overall_22, study_19_22$potential_19),4)
  ## Bias
bias_15_18 <- round(bias(study_15_18$overall_18, study_15_18$potential_15),4)
bias_16_19 <- round(bias(study_16_19$overall_19, study_16_19$potential_16),4)
bias_17_20 <- round(bias(study_17_20$overall_20, study_17_20$potential_17),4)
bias_18_21 <- round(bias(study_18_21$overall_21, study_18_21$potential_18),4)
bias_19_22 <- round(bias(study_19_22$overall_22, study_19_22$potential_19),4)

comparison <- c("15_18", "16_19", "17_20", "18_21", "19_22")
R2 <- c(rsq_15_18, rsq_16_19, rsq_17_20, rsq_18_21, rsq_19_22)
MAE <- c(mae_15_18, mae_16_19, mae_17_20, mae_18_21, mae_19_22)
bias <- c(bias_15_18, bias_16_19, bias_17_20, bias_18_21, bias_19_22)
  
data.frame(comparison, R2, MAE, bias)

thank you,

CodePudding user response：

So you have two lists of studies that you're comparing. Put them in lists:

study_overall <- list(study_15_18$overall_18, ...) # fill in ... as needed
study_potential <- list(study_15_18$potential_15, ...)

Now you can process those lists in parallel:

library(purrr)
cors <- map2_dbl(study_overall, study_potential, 
                  \(x, y) round(cor(x, y, method = "pearson"))
                )

Now you can put the resulting vectors into your data frame.