You can use the Enumerable#partition
method to split an enumerable object (like a hash) based on a condition. For example, to separate positive/negative values:
input.partition { |_, v| v.to_i < 0 }
# => [[["b", "-195"], ["c", "-163"]],
# [["a", "440"], ["d", "100"]]]
Then, to get the desired result, you can use map
and to_h
to convert the key/value arrays to hashes:
negative, positive = input.partition { |_, v| v.to_i < 0 }.map(&:to_h)
positive
# => {"a"=>"440", "d"=>"100"}
negative
# => {"b"=>"-195", "c"=>"-163"}
If you use a version of Ruby prior 2.1 you can replace the Array#to_h
method (that was introduced in Ruby 2.1) like this:
evens, odds = input.partition { |_, v| v.to_i.even? }
.map { |alist| Hash[alist] }