group_by and partition

Part of the Logic & Flow section of Coddy's Ruby journey — lesson 25 of 56.

Sometimes you want to split a collection rather than filter it. Two methods cover most cases.

group_by takes a block and returns a hash. Each unique block result becomes a key, mapping to an array of the elements that produced it:

words = ["hi", "hello", "yo", "hey", "bonjour"]
by_length = words.group_by { |w| w.length }
puts by_length.inspect
# {2=>["hi", "yo"], 5=>["hello"], 3=>["hey"], 7=>["bonjour"]}

It's the cleanest way to bucket items in one pass.

partition is the binary version: it returns a pair of arrays, the elements where the block was true, and the rest:

evens, odds = [1, 2, 3, 4, 5].partition { |n| n.even? }
puts evens.inspect   # [2, 4]
puts odds.inspect    # [1, 3, 5]

Use partition when there are exactly two buckets, group_by when there can be many.

Challenge

Medium

Read a comma-separated list of name:score entries (e.g. Alice:88,Bob:42,Charlie:75). Use group_by followed by partition to produce a small report.

Steps:

Parse each entry into a [name, score] pair (score is an integer)
group_by the pair's first letter of name (uppercase)
For each letter group, in alphabetical order, partition the entries into passing (score >= 60) and failing, then print one line per group:

<LETTER>: pass=<names_passing_joined_by_comma> fail=<names_failing_joined_by_comma>

Within each group, names appear in their original order. If a side has no names, leave it empty (e.g. fail=).

For input Alice:88,Adam:42,Bob:75,Beth:55,Cara:90, the output is:

A: pass=Alice fail=Adam
B: pass=Bob fail=Beth
C: pass=Cara fail=

Cheat sheet

group_by splits a collection into a hash keyed by block result:

words = ["hi", "hello", "yo"]
by_length = words.group_by { |w| w.length }
# {2=>["hi", "yo"], 5=>["hello"]}

partition splits into exactly two arrays (matching and non-matching):

evens, odds = [1, 2, 3, 4, 5].partition { |n| n.even? }
# evens => [2, 4], odds => [1, 3, 5]

Use partition for two buckets, group_by for many.

Try it yourself

entries = gets.chomp.split(",").map do |s|
  name, score = s.split(":")
  [name, score.to_i]
end

# TODO: group by first letter (upper), partition each group by score >= 60,
#       print 'X: pass=... fail=...' in alphabetical letter order