CODEWITHSUNDEEP
r
ahmad
ahmad

Reputation: 31

how to loop over all KPI's in my dataframe

for (i in colnames(outlier_df)){
  outliers <- outlier_df %>%
    group_by(factor, segment) %>%
    mutate(hinge_spread = 1.5*IQR(i),
           lwr = quantile(i, .25) - hinge_spread,
           upr = quantile(i, .75) + hinge_spread) %>%
    filter(i > upr | i < lwr)}

I am trying to loop over all kpi's in the dataframe and get the outlier the function works for one outlier but not in the loop

 Error in eval(substitute(expr), envir, enclos) : 
  missing values and NaN's not allowed if 'na.rm' is FALSE In addition: Warning message:
In quantile(as.numeric(x), c(0.25, 0.75), na.rm = na.rm, names = FALSE,  :
  NAs introduced by coercion

Upvotes: 1

Views: 33

Answers (1)

akrun
akrun

Reputation: 887741

The column names could be evaluated (!!) after converting to symbol or use .data. Also, instead of updating the same object in each iteration, assign the output to a list

outliers_list <- vector('list', ncol(outlier_df))
names(outliers_list) <- names(outlier_df)


for (i in colnames(outlier_df)){
  outliers_list[[i]] <- outlier_df %>%
    group_by(factor, segment) %>%
    mutate(hinge_spread = 1.5*IQR(.data[[i]]),
           lwr = quantile(.data[[i]], .25) - hinge_spread,
           upr = quantile(.data[[i]], .75) + hinge_spread) %>%
    ungroup %>%
    filter(.data[[i]] > upr | .data[[i]] < lwr)
 
}

Upvotes: 1

Related Questions