/* 

  Thermostat problem in Picat.

  OMEGA: Fast, casual Inference from Simple Parts 
  From  https://www.youtube.com/watch?v=oCvbqKE2tWA
  @ ~ 21min
  
  The Omega (Julia) model is here:
  https://github.com/zenna/Omega.jl/blob/master/docs/src/causal.md


  Cf my Gamble model gamble_.rkt

  This program was created by Hakan Kjellerstrand, hakank@gmail.com
  See also my Picat page: http://www.hakank.org/picat/

*/

import ppl_distributions, ppl_utils.
import util.

main => go.

/*

  var : is AC on?
  Probabilities:
  true: 1.0000000000000000
  mean = [true = 1.0]

  var : is window open?
  Probabilities:
  false: 1.0000000000000000
  mean = [false = 1.0]

  var : outside temp
  Probabilities (truncated):
  39.32539957103198: 0.0004081632653061
  38.045300216817118: 0.0004081632653061
  37.770576434195377: 0.0004081632653061
  37.326605503932697: 0.0004081632653061
  .........
  5.010186561945925: 0.0004081632653061
  4.625547895970008: 0.0004081632653061
  4.515905104009026: 0.0004081632653061
  4.024772753366593: 0.0004081632653061
  mean = 20.5763

  var : room temp
  Probabilities (truncated):
  25.781707100948193: 0.0004081632653061
  25.767139165653372: 0.0004081632653061
  25.600931029415307: 0.0004081632653061
  25.546328472361857: 0.0004081632653061
  .........
  14.134149911853081: 0.0004081632653061
  13.9584687654355: 0.0004081632653061
  13.564962657844635: 0.0004081632653061
  13.522571810027385: 0.0004081632653061
  mean = 19.9746

  var : thermostat
  Probabilities (truncated):
  25.781707100948193: 0.0004081632653061
  25.767139165653372: 0.0004081632653061
  25.600931029415307: 0.0004081632653061
  25.546328472361857: 0.0004081632653061
  .........
  14.134149911853081: 0.0004081632653061
  13.9584687654355: 0.0004081632653061
  13.564962657844635: 0.0004081632653061
  13.522571810027385: 0.0004081632653061
  mean = 19.9746

  var : this time
  Probabilities:
  morning: 0.3510204081632653
  evening: 0.3302040816326531
  afternoon: 0.3187755102040816
  mean = [morning = 0.35102,evening = 0.330204,afternoon = 0.318776]

*/
go ?=>
  reset_store,
  run_model(10_000,$model,[show_probs_trunc,mean]),
  nl,
  % fail,
  nl.
go => true.
  
  
model() =>
  Times = [morning,afternoon,evening],
  ThisTime = uniform_draw(Times),

  IsWindowOpen = flip(0.5),
  
  % a.c. is off when window is closed
  IsACOn = condt(IsWindowOpen,false,flip(0.5)),

  

  % hottest at noon, cool at night
  OutsideTemp = cases([ [ThisTime == morning,normal_dist(20,2)],
                        [ThisTime == afternoon,normal_dist(32,2)],
                        [ThisTime == evening,normal_dist(10,2)],
                        [true,normal_dist(25,2)]
                      ]),
  % a.c. chills the room
  RoomTemp = condt(IsACOn,normal_dist(20,2), normal_dist(25,2)),

  % great insulation
  Thermostat = condt(IsWindowOpen,(OutsideTemp + RoomTemp)/2,RoomTemp),

  % Observations
  observe(IsACOn),
  % observe(not IsWindowOpen),
  % observe(abs(RoomTemp-20.0)<1.0),
  % observe(RoomTemp > 20.0),
  % observe(ThisTime == evening),
  % observe(OutsideTemp < 10.0),
  % observe(Thermostat > 30.0),  
  
  if observed_ok then
    add_all([ ["this time",ThisTime],
              ["is AC on?",IsACOn],
              ["is window open?",IsWindowOpen],
              ["outside temp",OutsideTemp],
              ["room temp",RoomTemp],
              ["thermostat",Thermostat]
             ])
  end.