Clever Geek Handbook

Which of the following is most effective

If I would like to print x from 0 to 5

6.times {|x| p x}

(0..5).each {|x| p x}

0.upto(5) {|x| p x}

for x in 0..5
  p x
end
+3


source to share


2 answers


testmark / ips is the best tool for this.

require 'benchmark/ips'

Benchmark.ips do |x|
  x.report("times") { 6.times {|x| } }
  x.report("range iteration") { (0..5).each {|x| } }
  x.report("upto") { 0.upto(5) {|x| } }
  x.report("for") do
    for x in 0..5
    end
  end
end

Results in Ruby 2.2.0:



Calculating -------------------------------------
               times    88.567k i/100ms
     range iteration    88.519k i/100ms
                upto    86.749k i/100ms
                 for    84.118k i/100ms
-------------------------------------------------
               times      2.093M (± 0.2%) i/s -     10.451M
     range iteration      2.053M (± 0.4%) i/s -     10.268M
                upto      2.103M (± 0.2%) i/s -     10.583M
                 for      1.949M (± 0.2%) i/s -      9.758M

An empty body loop is the best idea here, as the I / O out time p

will likely overshadow the loop iteration times. The result is close enough not to matter.

+5


source


You will need to use something like Benchmark

The Benchmark module provides methods for measuring and reporting the time taken to run Ruby code.

require 'benchmark'
puts Benchmark.measure {6.times {|x| p x}}
puts Benchmark.measure {(0..5).each {|x| p x}}
puts Benchmark.measure {0.upto(5) {|x| p x}}

and you will see a result that looks like



1 ) 0.000000   0.000000   0.000000 (  0.000059)
1 ) 0.000000   0.000000   0.000000 (  0.000050)
3 ) 0.000000   0.000000   0.000000 (  0.000046)

now you can read more about benchmark here

you need to try a few times with 1,000,000. To get a good sensor, now the results will change with each calculation and with what else is running on that computer during the test

+2


source






More articles:


All Articles