Calculate rank of hash value and insert it into a CSV file in Ruby

Question

I have one hash

h1 = {"Cust1"=>500, "Cust4"=>400, "Cust2"=>100, "Cust3"=>100}

I want to insert this hash into a CSV file with the ranking of keys according to their value.

The sample output should look like this

ID,Sales,Rank Cust1,500,1 Cust4,400,2 Cust2,100,3 Cust3,100,3

I have to write the program in Ruby.

Answer 1

You can use each_with_index method of hash

arr = [['ID', 'Sales', 'Rank']]

h1.each_with_index do |(key, value), index|
  arr << [key, value, index + 1]
end

arr #[["ID", "Sales", "Rank"], ["Cust1", 500, 1], ["Cust4", 400, 2], ["Cust2", 100, 3], ["Cust3", 100, 4]]

Code for CSV

h1 = {"Cust1"=>500, "Cust4"=>400, "Cust2"=>100, "Cust3"=>100}
# You can skip following `sort_by` line if your hash is already ordered by desc value
h1 = h1.sort_by {|_k,v| v}.reverse.to_h
h2 = h1.group_by {|k,v| v }
require 'csv'
CSV.open("myfile.csv", "w") do |csv|
  csv << ['ID', 'Sales', 'Rank']
  h2.each_with_index do |(key, values), index|
    values.each do |value|
      csv << [value[0], key, index + 1]
    end
  end
end

Answer 2

The OP asks to calculate the rank; since 1.9 Ruby retains the insertion sequence of the hash, and in this instance, hash h1 is already in the correct order. But if the data came from a REST call or was not already sorted in ranked sequence the solution above would produce incorrect output. To fix this

require 'csv'

h1 = {"Cust1"=>500, "Cust2"=>100, "Cust4"=>400, "Cust3"=>100}

ranked_customers = h1.sort_by { |k, v| [-v, k] }.each_with_index { |e, i| e.push i + 1 }
ranked_customers.unshift %w[ID Sales Rank]

CSV { |csv_out| ranked_customers.each { |e| csv_out << e } }

produces

ID,Sales,Rank
Cust1,500,1
Cust4,400,2
Cust2,100,3
Cust3,100,4

Cust2 and Cust3 both have the same sales value, but Cust2 sorts ahead of Cust3 alphabetically.

Answer 3

I assume the key-value pairs of h1 are not necessarily in order of decreasing values.

We need to make two calculations to construct the CSV file.

order = h1.sort_by { |_,v| -v }
  #=> [["Cust1", 500], ["Cust4", 400], ["Cust2", 100], ["Cust3", 100]]
nbr_ranks = order.map(&:last).uniq.size
  #=> 3

The file can be constructed as follows.

require 'csv'

CSV.open('t.csv', 'w') do |csv|
  csv << ['ID', 'Sales', 'Rank']
  last_val = order.first.last
  order.each do |name, val|
    unless val == last_val
      nbr_ranks -= 1
      last_val = val
    end
    csv << [name, val, nbr_ranks]
  end
end

puts File.read('t.csv')
ID,Sales,Rank
Cust1,500,1
Cust4,400,2
Cust2,100,3
Cust3,100,4

---

If the key-value pairs of h1 are guaranteed to be in order of decreasing values (as in the example), there is no need to compute the array order.

nbr_ranks = order.map(&:last).uniq.size
  #=> 3

CSV.open('t.csv', 'w') do |csv|
  csv << ['ID', 'Sales', 'Rank']
  last_val = h1[h1.keys.first]
  h1.each do |name, val|
    unless val == last_val
      nbr_ranks -= 1
      last_val = val
    end
    csv << [name, val, nbr_ranks]
  end
end

Note:

last_val = h1[h1.keys.first]
  #=> 500