CODEWITHSUNDEEP
ruby
JVK
JVK

Reputation: 3912

Ruby iterating through thousands of records and resetting values (optimal solution)

mydata = {:data => [
        {
        :first_name = > "abc",
        :foo_id = > "21sd",
        :roll = > 43
        }, 
        {:first_name = > "def",
        :foo_id = > "2uf",
        :roll = > 81
        }, 
        {:first_name = > "xyz",
        :foo_id = > "ac32",
        :roll = > 2
        }
    ]
}

In mydata :roll has somehow corrupted value

I have a sorted hash, this has data as :foo_id => :roll

sorted = {"21sd" => 7, "ac32" => 89, "2uf" => 92}

(that is by sorted ascending correct value of :roll)

And I want to use this 'sorted' hash to rearrange 'mydata' and also over-ride the value of :roll in 'mydata' with the correct value from 'sorted' hash. So finally 'mydata' will look like

mydata = {:data => [
        {
        :first_name = > "abc",
        :foo_id = > "21sd",
        :roll = > 7
        }, 
        {:first_name = > "xyz",
        :foo_id = > "ac32",
        :roll = > 89
        },
        {:first_name = > "def",
        :foo_id = > "2uf",
        :roll = > 92
        }
    ]
}

UPADTE: :roll may not be unique in sorted

Consider mydata may have 100,000 hashes

I have achieved the result using nested loop of 'sorted' and in each iteration, searching the foo_id from 'mydata' and correcting the value and stroring the sorted data in new variable. Which is ugly.

correct = []
sorted.each {|k, v|
    mydata[:data].each {|h| # hate looping here
        if h[:foo_id] == k  # hate searching here, if i have 100,000 record in 'mydata'
            h[:roll] = v
           correct << h 
        end
    }
}
mydata = {:data => correct}

This is not an optimal solution if mydata contains large numbers of data sets. Anyone suggest some optimal solution?

Upvotes: 0

Views: 98

Answers (2)

Ismael
Ismael

Reputation: 16730

data =  [
        {
        :first_name => "abc",
        :foo_id => "21sd",
        :roll => 7
        },
        {:first_name => "xyz",
        :foo_id => "ac32",
        :roll => 89
        },
        {:first_name => "def",
        :foo_id => "2uf",
        :roll => 92
        }
    ]

sorted = {"21sd" => 7, "ac32" => 89, "2uf" => 92}

foo_id_name = data.map do |rec|
    { rec[:foo_id] =>  rec[:first_name] }
end.reduce(&:merge)

new_data = sorted.map do |foo_id, new_roll|
    {
        :first_name => foo_id_name[foo_id],
        :foo_id => foo_id,
        :roll => new_roll
    }
end

puts new_data.inspect #=> [{:first_name=>"abc", :foo_id=>"21sd", :roll=>7}, {:first_name=>"xyz", :foo_id=>"ac32", :roll=>89}, {:first_name=>"def", :foo_id=>"2uf", :roll=>92}]

Upvotes: 0

Philip Hallstrom
Philip Hallstrom

Reputation: 19899

This should work...

mydata[:data].each do |h|
  h[:roll] = sorted[h[:foo_id]]
end

There's no way to avoid looping through mydata[:data] though as it's an Array... but you don't need to stuff things into a new variable.. just update what you've got.

Unless I'm not understanding the problem...

EDIT: I wasn't picking up the sorting right. New solution:

mydata[:data].each do |h|
  h[:roll] = sorted[h[:foo_id]]
end
mydata[:data].sort_by!{|h| h[:roll]}

Upvotes: 2

Related Questions