1 function sheet09
   2 
   3 fprintf('p-Wert für k = 3, n = 1000, pi0 = 99.9%% = %f\n', ...
   4         one_sided_binominal(3, 1000, 0.999));
   5 
   6 X = [1.311, 1.136, 1.81, 0.827, -0.173, 0.351, -1.949, 0.973, ...
   7      0.617, -0.091, -0.155, -0.581, 0.452, 0.879, 0.17];
   8 
   9 fprintf('Exakter p-Wert für Daten (1) = %f\n', ...
  10         one_sided_gaussian(X));
  11 
  12 fprintf('Boostrapped p-Wert für Daten (1), N = 9999 = %f\n', ...
  13         one_sided_gaussian_bootstrap(X, 9999));
  14 
  15 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
  16 % Your solutions below
  17 
  18 % 1. Compute the p-value for a one-sided bionminal test.
  19 %
  20 % Input:
  21 %   k: observed incidents
  22 %   n: total number of samples
  23 %   pi0: probability under H_0
  24 %
  25 % Note: you may use "binopdf" or "binocdf", but no built in test
  26 % functions.
  27 function P = one_sided_binominal(k, n, pi0)
  28 % ...
  29 
  30 % 2. Compute the exact one-sided Gaussian test for
  31 %    mean 0 and variance 0.
  32 %
  33 % Input:
  34 %   X: sample realizations
  35 %
  36 % Note: you may use "normcdf", but no built in test functions.
  37 function P = one_sided_gaussian(X)
  38 % ...
  39 
  40 % 3. Compute a bootstrap estimate as in "Resamplingschema 2.1"
  41 %
  42 % Input:
  43 %    X: sample realizations
  44 %    B: number of boostrap samples
  45 function P = one_sided_gaussian_bootstrap(X, B)
  46 % ...