Idiomatic R often uses a neat syntactic sugar called shadowing.<\/p>\n\n\n\n
Imagine you have a dataframe (df) containing the costs of summer camps:<\/p>\n\n\n\n A common access pattern is to filter the data frame using the column names. Note however that we pass an expression – not a lambda, as we would in other languages:<\/p>\n\n\n\n To learn why this works, we can implement a crude version of a filtering function ourselves.<\/p>\n\n\n\n First, let’s define a function that takes a dataframe and an expression:<\/p>\n\n\n\n If you call this as “filter(df, Week == 1)”, it will print out:<\/p>\n\n\n\n This output is the .toString equivalent for the abstract syntax tree of the expression provided in Note that if we replace There is no variable named ‘a’ in scope in the environment. We only get this error when the expression is used, as the arguments are promises that represent the result of the expression passed in, and are lazily evaluated.<\/p>\n\n\n\n The difference between the expression To complete our Then all we need to do is to loop over the rows in the dataframe, populating the variables as we look at each row. The For our final step, we can put this all together into a function which prints out rows that match our expression. This adds a call to And there you have it!<\/p>\n","protected":false},"excerpt":{"rendered":" Idiomatic R often uses a neat syntactic sugar called shadowing. Imagine you have a dataframe (df) containing the costs of summer camps: Child Week Cost 1 1 $60 1 2 $200 2 1 $400 2 2 $275 2 2 $1000 A common access pattern is to filter the data frame using the column names. Note … <\/p>\nChild<\/td> Week<\/td> Cost<\/td><\/tr> 1<\/td> 1<\/td> $60<\/td><\/tr> 1<\/td> 2<\/td> $200<\/td><\/tr> 2<\/td> 1<\/td> $400<\/td><\/tr> 2<\/td> 2<\/td> $275<\/td><\/tr> 2<\/td> 2<\/td> $1000<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n df[Week == 1]<\/code><\/pre>\n\n\n\nsimple_filter <- function(df, e) {\n print(enexpr(e))\n}<\/code><\/pre>\n\n\n\nWeek == 1<\/code><\/code><\/pre>\n\n\n\ne<\/code>. <\/p>\n\n\n\nprint(enexpr(e))<\/code> with print(e)<\/code>, we’ll get an error:<\/p>\n\n\n\nError in print(e) : object 'Week' not found<\/code><\/p>\n\n\n\nWeek == <\/code>1 and the lambda (Week) => Week == 1<\/code> is that in a lambda, context is provided. Context includes both arguments and variables the lambda closes over. These form the “environment” in which the function runs.<\/p>\n\n\n\nfilter<\/code> implementation, we need to build an environment in which to evaluate the variables:<\/p>\n\n\n\nenv <- new.env()<\/code><\/code><\/pre>\n\n\n\nassign<\/code> function inserts a value into the environment:<\/p>\n\n\n\ncols <- colnames(df)\nfor(i in 1:nrow(df)) { \n for(j in 1:ncol(df)) {\n col <- cols[j]\n assign(col, df[i, j], env)\n }\n\n ...\n}<\/code><\/pre>\n\n\n\neval<\/code>, which evaluates the given expression in the context of the environment we’ve built from the row:<\/p>\n\n\n\nsimple_filter <- function(df, e) {\n query <- enexpr(e)\n cols <- colnames(df)\n env <- new.env()\n for(i in 1:nrow(df)) { \n for(j in 1:ncol(df)) {\n col <- cols[j]\n assign(col, df[i, j], env)\n }\n\n if(eval(query, envir = env)) {\n print(df[i, ])\n }\n }\n}<\/code><\/pre>\n\n\n\n