mutate, transmute, morph

[moodymudskipper]

In dplyr :

• mutate() creates or modifies columns and keep existing cols
• transmute() creates or modifies columns and remove others
• morph() (will) create or modify columns, and remove columns used in the call

In data.table

• := mutates in place, allowing a flexible lhs :
  • use foo <- "bar"; dt[, (foo) := expr] to create a bar column
  • zeallot like syntax to do dt[, list(var1, var2) := something]
• .() transmutes, not in place (or summarizes, depending if by argument.

In tb :

• We can transmute in .j
• We can mutate in ...
• We can morph in both by putting the lhs inside double parentheses, which can be used to rename
• multi assign is supported