task3 solution

Author: proton

week3/quicksort/interview-questions/task3/solution.rb

@@ -1,25 +1,60 @@
+class Array
+  def swap(i, j)
+    return if i == j
+    self[i], self[j] = self[j], self[i]
+  end
+end
 
+### Decimal dominants
+# Given an array with `n` keys, design an algorithm to find all values that occur more than `n/10` times. The expected running time of your algorithm should be linear.
 
+n = 1000
 
+a = (n / 2).times.map { rand(0..4) } + (n / 2).times.map { rand(0..99) }
+d = 10
 
+expected = a.tally.select { |_, v| v > n / d }.keys
 
+### Solution
 
-----------------------------------
+def count_dominants(arr, lo, hi, d, result)
+  return if lo >= hi
 
-Hint: determine the
-(
-�
-/
-10
-)
-�
-ℎ
-(n/10)
-th
-  largest key using quickselect and check if it occurs more than
-�
-/
-10
-n/10 times.
+  i  = lo
+  lt = lo
+  rt = hi
+  pivot = arr[lo]
 
-Alternate solution hint: use 9 counters.
+  while i <= rt
+    case arr[i] <=> pivot
+    when -1
+      arr.swap(i, lt)
+      i  += 1
+      lt += 1
+    when 1
+      arr.swap(i, rt)
+      rt -= 1
+    else
+      i += 1
+    end
+  end
+
+  count = i - lt
+
+  result << arr[lt] if count > arr.size / d
+
+  count_dominants(arr, lo,     lt - 1, d, result)
+  count_dominants(arr, rt + 1, hi,     d, result)
+end
+
+result = []
+count_dominants(a, 0, a.size - 1, d, result)
+
+p expected.sort!
+p result.sort!
+p expected == result
+
+
+# Hint: determine the `(n/10)^th` largest key using quickselect and check if it occurs more than `n/10` times.
+
+# Alternate solution hint: use 9 counters.