outline
basic parameters
simulating fails per month
sampling from the prior distribution
samping from the posterior using importance sampling
computing weights

s =

   1.2487e-17

sort and display results
true -- lower - expected - upper for lambda1

ans =

    0.9000    0.6780    0.8668    0.9961

true -- lower - expected - upper for lambda2

ans =

    0.1000    0.0248    0.1220    0.3705

true -- lower - expected - upper for alpha

ans =

    0.1000    0.0935    0.1119    0.1915

posterior predictive
#failed in month I+1, #predicted, 2*SD

ans =

  925.0000  877.3214   51.6595

cross validation

i =

     8

sampling from prior (again)
compute the weights for these prior samples
computing weights

s =

   1.7877e-15

cross val values

x_post_sim_cross =

    0.8421    0.1601    0.1271

obs(c) - expected value 2*SD

ans =

  929.0000  887.9179   50.0311


i =

     9

sampling from prior (again)
compute the weights for these prior samples
computing weights

s =

   1.2785e-15

cross val values

x_post_sim_cross =

    0.8153    0.1804    0.1322

obs(c) - expected value 2*SD

ans =

  908.0000  883.1829   50.7303


i =

    10

sampling from prior (again)
compute the weights for these prior samples
computing weights

s =

   5.8374e-16

cross val values

x_post_sim_cross =

    0.6677    0.2955    0.1658

obs(c) - expected value 2*SD

ans =

  870.0000  892.3214   51.6021