Get a basic benchmark runner going.

benchmark/fib.lua

@@ -3,4 +3,9 @@ function fib(n)
   return fib(n - 2) + fib(n - 1)
 end
 
-io.write(fib(35) .. "\n")
+local start = os.clock()
+for i = 1, 5 do
+  io.write(fib(30) .. "\n")
+end
+local elapsed = os.clock() - start
+io.write(string.format("elapsed: %.8f\n", elapsed))

benchmark/fib.py

@@ -1,5 +1,10 @@
+import time
+
 def fib(n):
   if n < 2: return n
   return fib(n - 1) + fib(n - 2)
 
-print fib(35)
+start = time.clock()
+for i in range(0, 5):
+  print(fib(30))
+print("elapsed: " + str(time.clock() - start))

benchmark/fib.rb

@@ -0,0 +1,13 @@
+def fib(n)
+  if n < 2 then
+    n
+  else
+    fib(n - 1) + fib(n - 2)
+  end
+end
+
+start = Time.now
+for i in 0...5
+  puts fib(30)
+end
+puts "elapsed: " + (Time.now - start).to_s

benchmark/fib.wren

@@ -6,4 +6,10 @@ var fib = fn(n) {
   }
 }
 
-io.write(fib.call(35))
+var start = OS.clock
+var i = 0
+while (i < 5) {
+  io.write(fib.call(30))
+  i = i + 1
+}
+io.write("elapsed: " + (OS.clock - start).toString)

benchmark/run_all

@@ -0,0 +1,65 @@
+#!/usr/bin/python
+
+import math
+import re
+import subprocess
+import sys
+
+FIB_OUTPUT_PATTERN = re.compile(r"""832040
+832040
+832040
+832040
+832040
+elapsed: (\d+\.\d+)""", re.MULTILINE)
+
+BENCHMARKS = [
+  ("fib", FIB_OUTPUT_PATTERN)
+]
+
+LANGUAGES = [
+  ("wren",   "../build/Release/wren", ".wren"),
+  ("lua",    "lua",                   ".lua"),
+  ("python", "python",                ".py"),
+  ("ruby",   "ruby",                  ".rb")
+]
+
+def calc_stats(nums):
+  """Calculates the mean, median, and std deviation of a list of numbers."""
+  mean = sum(nums) / len(nums)
+  nums.sort()
+  median = nums[(len(nums) - 1) / 2]
+  diffs = ((n - mean) * (n - mean) for n in nums)
+  std_dev = math.sqrt(sum(diffs) / len(nums))
+  return [mean, median, std_dev]
+
+
+def run_benchmark_once(benchmark, language):
+  args = [language[1], benchmark[0] + language[2]]
+  out = subprocess.check_output(args, universal_newlines=True)
+  match = benchmark[1].match(out)
+  if match:
+    return float(match.group(1))
+  else:
+    return None
+
+
+def run_benchmark(benchmark, language):
+  print "{0} - {1:10s}".format(benchmark[0], language[0]),
+
+  times = []
+  for i in range(0, 7):
+    times.append(run_benchmark_once(benchmark, language))
+    sys.stdout.write(".")
+
+  if None in times:
+    print "error"
+    return
+
+  times.sort()
+  stats = calc_stats(times)
+  print " mean: {0:.4f} median: {1:.4f} std_dev: {2:.4f}".format(
+      stats[0], stats[1], stats[2])
+
+for benchmark in BENCHMARKS:
+  for language in LANGUAGES:
+    run_benchmark(benchmark, language)