test_decision_stump.py

import decision_stump

import random
import numpy

def shuffle(ary):
  a = len(ary)
  b = a-1
  for d in range(b,0,-1):
    e = random.randint(0,d)
    if e == d:
      continue
    ary[d],ary[e] = ary[e],ary[d]
  return ary

def test_build_stump_1d():
  x = [0, 1, 2, 4]
  y = [-1, -1, 1, 1]
  w = [0.25, 0.25, 0.25, 0.25]
  stump = decision_stump.build_stump_1d(x, y, w)
  assert(stump.threshold == 1.5)
  assert(stump.s == 1)
  assert(stump.err == 0)
  x = [0, 1, 2, 4]
  y = [1, 1, -1, -1]
  stump = decision_stump.build_stump_1d(x, y, w)
  assert(stump.threshold == 1.5)
  assert(stump.s == -1)
  assert(stump.err == 0)

def test_build_stump_1d_shuffle():
  N = 100
  x = numpy.random.rand(N)
  y = numpy.random.rand(N)
  y[numpy.where(y < 0.5)] = -1
  y[numpy.where(y > 0.5)] = 1
  w = numpy.ones(N)/N
  stump = decision_stump.build_stump_1d(x, y, w)
  ind = shuffle(range(0, N))
  stump_shuffled = decision_stump.build_stump_1d(x[ind], y[ind], w[ind])
  assert(stump == stump_shuffled)
  ind = shuffle(range(0, N))
  yy = y.copy()
  for i in range(N):
    yy[i] = -y[ind[i]]
  stump_shuffled = decision_stump.build_stump_1d(x[ind], yy, w[ind])
  assert(stump.err == stump_shuffled.err)
  assert(stump.threshold == stump_shuffled.threshold)
  assert(stump.s == -1.0*stump_shuffled.s)