In mkuhn/dict: Dictionary with arbitrary keys and values

Introduction to the dict package

The dict package exports two functions: dict() and numvecdict(). The former creates a Python-style dictionary (or hash table) that can take any R object as value, and numbers and strings (and the respective vectors) as keys. The latter behaves like a Python defaultdict(list), but takes only numeric vectors as values.

Overview of member functions: dict

| function | description | |--------------------------|-----------------------------------------| | d[[key]] <- value | assignment | | d[[key]] | retrieval, error if key not present | | d$get(key) | retrieval, return NULL if not present | | d$get(key, "default") | retrieval, return second parameter if not present | | d$keys() | list of keys | | d$values() | list of values | | d$items() | list of items (as a list of lists) | | d$length() | number of items |

Overview of member functions: numvdict

| function | description | |--------------------------|-----------------------------------------| | d[[key]] <- value | assignment (must be number or numeric vector) | | d[[key]] | retrieval, return empty vector if not present | | d$get(key) | retrieval, return empty vector if not present | | d$get(key, NA) | retrieval, return second parameter if not present | | d$append_number(key, x)| append a number to an entry (creates a new entry if necessary) | | d$append_items(d2) | append items from another dictionary | | d$each_mean() (also: median, max, min) | return new dictionary with the mean of each value | | d$inplace_mean() (also: median, max, min) | convert values to their mean for this dictionary | | d$keys() | list of keys | | d$values() | list of values | | d$items() | list of items (as a list of lists) | | d$length() | number of items |

Usage examples: dict

To get a new, empty dictionary, call the dict function, optionally with argumebts for initialization:

library(dict)

d <- dict( list("Initial", "keys"), list("Initial", "values") )

d <- dict()

You can then assign items using numbers and strings (and vectors of these) as keys:

d[[1]] <- "First entry!"
d[[c(2,3)]] <- 5
d[["a string"]] <- list(A = 1, B = 2)
d[[c("A", "B")]] <- function(x) 1+x

...and then retrieve them again:

d[[c(2,3)]]
f <- d[[c("A", "B")]]
f(41)

Accessing a non-existing key leads to an error:

d[["?"]]

However, you can use d$get to return NA or another default value:

d$get("?")
d$get("?", "default")

You can access the keys, values, items, and length:

d$keys()
d$values()
d$items()
d$length()

Usage examples: numvecdict

To get a new, empty dictionary, call the numvecdict function:

d <- numvecdict()

This dictionary only contains vectors of numbers, which you can assign:

d[[1]] <- 1
d[[2]] <- c(2,3)
d[["a string"]] <- 1:10

You can append numbers to the lists. If you specify a new key, the entry will be generated automatically:

d$append_number(1, 2)
d$append_number(2, 3)
d$append_number("answers", 42)

Then, you can retrieve the vectors you stored:

d$get(1)
d[[2]]
d[["answers"]]
d[["completely new and never seen key"]]

It is possible to add the values of one numvecdict to the other:

a <- numvecdict()
b <- numvecdict()

a[[1]] <- 1
b[[1]] <- 2
b[[2]] <- 3

a$append_items(b)

a$items()

You can extract the means into a new dictionary. The same works for the median, min, and max.

mean_d <- d$each_mean()
mean_d$items()

It is also possible to replace each value with its mean:

d$inplace_mean()
d$items()

mkuhn/dict documentation built on May 23, 2019, 2:03 a.m.