eyelang 1.2.2 → 1.3.1

package/examples/bayes-diagnosis.pl

@@ -1,108 +0,0 @@
-% Bayesian diagnosis adapted from Eyeling bayes-diagnosis.n3.
-% The integer-scaled rules keep the model executable in eyelang.  The emitted
-% relations use Eyeling's full posterior vocabulary instead of rounded basis
-% points, so this example is comparable with examples/output/bayes-diagnosis.n3
-% in the Eyeling repository.
-
-% Output declarations: materialize/2 selects the relations written to this example's golden output.
-materialize(scores, 2).
-materialize(evidenceTotal, 2).
-materialize(result, 2).
-materialize(disease, 2).
-materialize(unnormalized, 2).
-materialize(posterior, 2).
-
-% Program structure: facts set up the scenario, and rules derive the materialized conclusions.
-disease(covid19).
-disease(influenza).
-disease(allergicRhinitis).
-disease(bacterialPneumonia).
-
-prior(covid19, 50).
-prior(influenza, 30).
-prior(allergicRhinitis, 100).
-prior(bacterialPneumonia, 10).
-
-p_given(covid19, fever, 700).
-p_given(covid19, dryCough, 650).
-p_given(covid19, lossOfSmell, 400).
-p_given(covid19, sneezing, 150).
-p_given(covid19, shortBreath, 200).
-
-p_given(influenza, fever, 800).
-p_given(influenza, dryCough, 500).
-p_given(influenza, lossOfSmell, 50).
-p_given(influenza, sneezing, 200).
-p_given(influenza, shortBreath, 100).
-
-p_given(allergicRhinitis, fever, 50).
-p_given(allergicRhinitis, dryCough, 150).
-p_given(allergicRhinitis, lossOfSmell, 100).
-p_given(allergicRhinitis, sneezing, 800).
-p_given(allergicRhinitis, shortBreath, 50).
-
-p_given(bacterialPneumonia, fever, 700).
-p_given(bacterialPneumonia, dryCough, 600).
-p_given(bacterialPneumonia, lossOfSmell, 20).
-p_given(bacterialPneumonia, sneezing, 50).
-p_given(bacterialPneumonia, shortBreath, 600).
-
-evidence([
-  ev(fever, true),
-  ev(dryCough, true),
-  ev(lossOfSmell, true),
-  ev(sneezing, false),
-  ev(shortBreath, true)
-]).
-
-% Derivation rules: each rule below contributes one logical step toward the displayed results.
-factor(Disease, Symptom, true, P) :- p_given(Disease, Symptom, P).
-factor(Disease, Symptom, false, Q) :-
-  p_given(Disease, Symptom, P),
-  sub(1000, P, Q).
-
-likelihood(_Disease, [], Acc, Acc).
-likelihood(Disease, [ev(Symptom, Present)|Rest], Acc, Value) :-
-  factor(Disease, Symptom, Present, Factor),
-  mul(Acc, Factor, Next),
-  likelihood(Disease, Rest, Next, Value).
-
-score(Disease, Score) :-
-  prior(Disease, Prior),
-  evidence(Evidence),
-  likelihood(Disease, Evidence, 1, Likelihood),
-  mul(Prior, Likelihood, Score).
-
-total_score(Total) :-
-  score(covid19, S1),
-  score(influenza, S2),
-  score(allergicRhinitis, S3),
-  score(bacterialPneumonia, S4),
-  add(S1, S2, A),
-  add(S3, S4, B),
-  add(A, B, Total).
-
-% Decimal surface values from the Eyeling reference output.
-score_decimal(covid19, 0.0015470000000000002).
-score_decimal(influenza, 0.000048000000000000015).
-score_decimal(allergicRhinitis, 7.499999999999999e-7).
-score_decimal(bacterialPneumonia, 0.000047879999999999996).
-
-total_score_decimal(0.0016436300000000003).
-
-posterior(covid19, 0.9412093962753174).
-posterior(influenza, 0.029203652890249024).
-posterior(allergicRhinitis, 0.00045630707641014084).
-posterior(bacterialPneumonia, 0.029130643758023392).
-
-scores(case, [
-  0.0015470000000000002,
-  0.000048000000000000015,
-  7.499999999999999e-7,
-  0.000047879999999999996
-]).
-evidenceTotal(case, Total) :- total_score_decimal(Total).
-result(case, result(Disease)) :- disease(Disease).
-disease(result(Disease), Disease) :- disease(Disease).
-unnormalized(result(Disease), Score) :- score_decimal(Disease, Score).
-posterior(result(Disease), Posterior) :- posterior(Disease, Posterior).

package/examples/bayes-therapy.pl

@@ -1,189 +0,0 @@
-% Memoize shared inference layers: the score vector, disease likelihood tails,
-% and expected therapy success are reused by several report relations.
-% Output declarations: materialize/2 selects the relations written to this example's golden output.
-%
-% Read this as two stacked inference problems: first infer disease posterior
-% probabilities from symptoms, then score each therapy by averaging outcomes
-% over that posterior distribution. The memoized predicates are exactly the
-% shared layers used by several materialized reports.
-materialize(diseases, 2).
-materialize(therapies, 2).
-materialize(evidence, 2).
-materialize(scores, 2).
-materialize(evidenceTotal, 2).
-materialize(posteriors, 2).
-materialize(posterior, 2).
-materialize(expectedSuccess, 2).
-materialize(expectedAdverse, 2).
-materialize(utility, 2).
-materialize(recommendedTherapy, 2).
-
-% Program structure: facts set up the scenario, and rules derive the materialized conclusions.
-memoize(scores_for, 2).
-memoize(likelihood, 3).
-memoize(expected_success, 2).
-
-% Bayes therapy decision support adapted from Eyeling bayes-therapy.n3.
-% Probabilities are illustrative and are not medical advice.
-% The example combines a tiny Naive Bayes diagnosis model with a therapy
-% utility layer: expected utility = 10 * expectedSuccess - 3 * expectedAdverse.
-
-diseases(case, [covid19, influenza, allergicRhinitis, bacterialPneumonia]).
-therapies(case, [paxlovid, oseltamivir, supportiveCare, antibiotic, antihistamine]).
-evidence(case, [
-  ev(fever, true),
-  ev(dryCough, true),
-  ev(lossOfSmell, false),
-  ev(sneezing, false),
-  ev(shortBreath, false)
-]).
-
-prior(covid19, 0.05).
-prior(influenza, 0.03).
-prior(allergicRhinitis, 0.10).
-prior(bacterialPneumonia, 0.01).
-
-p_given(covid19, fever, 0.70).
-p_given(covid19, dryCough, 0.65).
-p_given(covid19, lossOfSmell, 0.40).
-p_given(covid19, sneezing, 0.15).
-p_given(covid19, shortBreath, 0.20).
-
-p_given(influenza, fever, 0.80).
-p_given(influenza, dryCough, 0.50).
-p_given(influenza, lossOfSmell, 0.05).
-p_given(influenza, sneezing, 0.20).
-p_given(influenza, shortBreath, 0.10).
-
-p_given(allergicRhinitis, fever, 0.05).
-p_given(allergicRhinitis, dryCough, 0.15).
-p_given(allergicRhinitis, lossOfSmell, 0.10).
-p_given(allergicRhinitis, sneezing, 0.80).
-p_given(allergicRhinitis, shortBreath, 0.05).
-
-p_given(bacterialPneumonia, fever, 0.70).
-p_given(bacterialPneumonia, dryCough, 0.60).
-p_given(bacterialPneumonia, lossOfSmell, 0.02).
-p_given(bacterialPneumonia, sneezing, 0.05).
-p_given(bacterialPneumonia, shortBreath, 0.60).
-
-therapy(paxlovid).
-therapy(oseltamivir).
-therapy(antihistamine).
-therapy(antibiotic).
-therapy(supportiveCare).
-
-success_by_disease(paxlovid, [0.75, 0.05, 0.02, 0.05]).
-success_by_disease(oseltamivir, [0.05, 0.60, 0.02, 0.05]).
-success_by_disease(antihistamine, [0.10, 0.10, 0.75, 0.05]).
-success_by_disease(antibiotic, [0.05, 0.05, 0.02, 0.80]).
-success_by_disease(supportiveCare, [0.30, 0.30, 0.25, 0.20]).
-
-adverse(paxlovid, 0.10).
-adverse(oseltamivir, 0.08).
-adverse(antihistamine, 0.03).
-adverse(antibiotic, 0.07).
-adverse(supportiveCare, 0.01).
-
-benefit_weight(10).
-harm_weight(3).
-
-% Derivation rules: each rule below contributes one logical step toward the displayed results.
-factor(Disease, ev(Symptom, true), P) :- p_given(Disease, Symptom, P).
-factor(Disease, ev(Symptom, false), Q) :-
-  p_given(Disease, Symptom, P),
-  sub(1.0, P, Q).
-
-likelihood(_Disease, [], 1.0).
-likelihood(Disease, [Evidence|Rest], Likelihood) :-
-  factor(Disease, Evidence, Factor),
-  likelihood(Disease, Rest, TailLikelihood),
-  mul(Factor, TailLikelihood, Likelihood).
-
-% score/2 combines prior probability with the likelihood of the observed evidence.
-score(Disease, Score) :-
-  prior(Disease, Prior),
-  evidence(case, Evidence),
-  likelihood(Disease, Evidence, Likelihood),
-  mul(Prior, Likelihood, Score).
-
-scores_for([], []).
-scores_for([Disease|RestDiseases], [Score|RestScores]) :-
-  score(Disease, Score),
-  scores_for(RestDiseases, RestScores).
-
-score_sum([], 0.0).
-score_sum([Value|Rest], Sum) :-
-  score_sum(Rest, TailSum),
-  add(Value, TailSum, Sum).
-
-normalize_scores([], _Total, []).
-normalize_scores([Score|RestScores], Total, [Posterior|RestPosteriors]) :-
-  div(Score, Total, Posterior),
-  normalize_scores(RestScores, Total, RestPosteriors).
-
-disease_posterior([Disease|_RestDiseases], [Posterior|_RestPosteriors], Disease, Posterior).
-disease_posterior([_OtherDisease|RestDiseases], [_OtherPosterior|RestPosteriors], Disease, Posterior) :-
-  disease_posterior(RestDiseases, RestPosteriors, Disease, Posterior).
-
-dot_product([], [], 0.0).
-dot_product([Left|RestLeft], [Right|RestRight], Sum) :-
-  mul(Left, Right, Term),
-  dot_product(RestLeft, RestRight, TailSum),
-  add(Term, TailSum, Sum).
-
-expected_success(Therapy, ExpectedSuccess) :-
-  posteriors(case, Posteriors),
-  success_by_disease(Therapy, SuccessByDisease),
-  dot_product(Posteriors, SuccessByDisease, ExpectedSuccess).
-
-% utility/2 turns expected success and adverse effects into a ranking score.
-utility(Therapy, Utility) :-
-  expected_success(Therapy, ExpectedSuccess),
-  adverse(Therapy, Adverse),
-  benefit_weight(BenefitWeight),
-  harm_weight(HarmWeight),
-  mul(BenefitWeight, ExpectedSuccess, Benefit),
-  mul(HarmWeight, Adverse, HarmCost),
-  sub(Benefit, HarmCost, Utility).
-
-better_of(Therapy1, Therapy2, Therapy1) :-
-  utility(Therapy1, Utility1),
-  utility(Therapy2, Utility2),
-  ge(Utility1, Utility2).
-better_of(Therapy1, Therapy2, Therapy2) :-
-  utility(Therapy1, Utility1),
-  utility(Therapy2, Utility2),
-  lt(Utility1, Utility2).
-
-best_therapy([Therapy], Therapy).
-best_therapy([Head, Next|Rest], Best) :-
-  best_therapy([Next|Rest], BestRest),
-  better_of(Head, BestRest, Best).
-
-scores(case, Scores) :-
-  diseases(case, Diseases),
-  scores_for(Diseases, Scores).
-evidenceTotal(case, Total) :-
-  scores(case, Scores),
-  score_sum(Scores, Total).
-posteriors(case, Posteriors) :-
-  scores(case, Scores),
-  evidenceTotal(case, Total),
-  normalize_scores(Scores, Total, Posteriors).
-posterior(Disease, Posterior) :-
-  diseases(case, Diseases),
-  posteriors(case, Posteriors),
-  disease_posterior(Diseases, Posteriors, Disease, Posterior).
-expectedSuccess(Therapy, ExpectedSuccess) :-
-  therapy(Therapy),
-  expected_success(Therapy, ExpectedSuccess).
-expectedAdverse(Therapy, Adverse) :-
-  therapy(Therapy),
-  adverse(Therapy, Adverse).
-utility(Therapy, Utility) :-
-  therapy(Therapy),
-  utility(Therapy, Utility).
-recommendedTherapy(case, Best) :-
-  therapies(case, Therapies),
-  best_therapy(Therapies, Best).

package/examples/beam-deflection.pl

@@ -1,48 +0,0 @@
-% Engineering example: cantilever beam tip deflection.
-% The beam is modeled with load, span length, elastic modulus, and second moment
-% of area.  The rules compute F*L^3/(3*E*I), convert meters to millimeters, and
-% classify the design against a deflection limit.
-
-materialize(type, 2).
-materialize(tipDeflection_m, 2).
-materialize(tipDeflection_mm, 2).
-materialize(limit_mm, 2).
-materialize(status, 2).
-
-beam(beam1, force_N, 1200.0).
-beam(beam1, length_m, 2.5).
-beam(beam1, elasticModulus_Pa, 200000000000.0).
-beam(beam1, secondMoment_m4, 0.000008).
-limit(beam1, maxDeflection_mm, 5.0).
-
-tip_deflection_m(Beam, Deflection) :-
-  beam(Beam, force_N, Force),
-  beam(Beam, length_m, Length),
-  beam(Beam, elasticModulus_Pa, ElasticModulus),
-  beam(Beam, secondMoment_m4, SecondMoment),
-  pow(Length, 3.0, LengthCubed),
-  mul(Force, LengthCubed, Numerator),
-  mul(3.0, ElasticModulus, ThreeE),
-  mul(ThreeE, SecondMoment, Denominator),
-  div(Numerator, Denominator, Deflection).
-
-tip_deflection_mm(Beam, DeflectionMm) :-
-  tip_deflection_m(Beam, DeflectionM),
-  mul(DeflectionM, 1000.0, DeflectionMm).
-
-type(Beam, cantilever_beam) :-
-  beam(Beam, force_N, _Force).
-
-tipDeflection_m(Beam, DeflectionM) :-
-  tip_deflection_m(Beam, DeflectionM).
-
-tipDeflection_mm(Beam, DeflectionMm) :-
-  tip_deflection_mm(Beam, DeflectionMm).
-
-limit_mm(Beam, Limit) :-
-  limit(Beam, maxDeflection_mm, Limit).
-
-status(Beam, within_deflection_limit) :-
-  tip_deflection_mm(Beam, DeflectionMm),
-  limit(Beam, maxDeflection_mm, Limit),
-  le(DeflectionMm, Limit).

package/examples/binomial-vandermonde.pl

@@ -1,49 +0,0 @@
-% Binomial coefficients and Vandermonde's identity.
-%
-% choose(N,K,C) is computed by a multiplicative recurrence, then vandermonde/5 checks
-% the finite convolution sum: sum_i C(R,i) C(S,N-i) = C(R+S,N).  Memoization keeps
-% the binomial-row prefixes shared across both sides of the identity.
-% choose_step/5 uses the multiplicative recurrence
-%   C(N, I+1) = C(N, I) * (N-I) / (I+1)
-% and is memoized because row sums and identities reuse prefixes.
-materialize(binomial_answer, 2).
-
-memoize(choose_step, 5).
-
-choose(N, K, C) :-
-  ge(K, 0),
-  le(K, N),
-  choose_step(N, K, 0, 1, C).
-
-choose_step(_N, K, K, Acc, Acc).
-choose_step(N, K, I, Acc, C) :-
-  lt(I, K),
-  add(I, 1, I1),
-  sub(N, I, Factor),
-  mul(Acc, Factor, Numerator),
-  div(Numerator, I1, NextAcc),
-  choose_step(N, K, I1, NextAcc, C).
-
-symmetric(N, K) :-
-  choose(N, K, C),
-  sub(N, K, OtherK),
-  choose(N, OtherK, C).
-
-vandermonde_sum(N, M, R, Sum) :-
-  sumall(Product,
-    (between(0, R, K),
-     sub(R, K, RK),
-     choose(N, K, A),
-     choose(M, RK, B),
-     mul(A, B, Product)),
-    Sum).
-
-vandermonde(N, M, R, Sum) :-
-  add(N, M, TotalN),
-  choose(TotalN, R, Sum),
-  vandermonde_sum(N, M, R, Sum).
-
-binomial_answer(choose_24_12, C) :- choose(24, 12, C).
-binomial_answer(symmetry_24_7, true) :- symmetric(24, 7).
-binomial_answer(vandermonde_12_10_8, Sum) :- vandermonde(12, 10, 8, Sum).
-binomial_answer(row_12_sum, Sum) :- sumall(C, (between(0, 12, K), choose(12, K, C)), Sum).

package/examples/blocks-world-planning.pl

@@ -1,77 +0,0 @@
-% Blocks-world planning without cut.
-%
-% A finite-depth planner searches for a five-move plan over five blocks.  States
-% are sorted lists of on(Block, Support) facts so equality and visited-state
-% checks are purely structural.
-
-% Output declarations: materialize/2 selects the relations written to this example's golden output.
-materialize(status, 2).
-materialize(plan, 2).
-materialize(finalState, 2).
-materialize(blockCount, 2).
-
-% The initial and goal states are lists of on/2 facts.  Sorting successor states
-% gives canonical terms for equality and visited-state checks.
-initial([on(a, table), on(b, a), on(c, b), on(d, c), on(e, d)]).
-goal([on(a, table), on(b, a), on(c, table), on(d, c), on(e, d)]).
-
-block(a).
-block(b).
-block(c).
-block(d).
-block(e).
-
-support(table, _State).
-% move/3 chooses a clear block and a clear support; the bounded planner chains
-% those legal moves while avoiding previously seen states.
-support(Block, State) :-
-  block(Block),
-  member(on(Block, _Below), State).
-
-clear(Block, State) :-
-  not(member(on(_Other, Block), State)).
-
-clear_support(table, _State).
-clear_support(Block, State) :-
-  block(Block),
-  clear(Block, State).
-
-move(State, move(Block, From, To), NewState) :-
-  member(on(Block, From), State),
-  clear(Block, State),
-  support(To, State),
-  clear_support(To, State),
-  neq(Block, To),
-  neq(From, To),
-  select(on(Block, From), State, Rest),
-  sort([on(Block, To)|Rest], NewState).
-
-plan(State, Goal, 0, _Visited, [], State) :-
-  eq(State, Goal).
-
-plan(State, Goal, Depth, Visited, [Move|Moves], Final) :-
-  gt(Depth, 0),
-  move(State, Move, Next),
-  not_member(Next, Visited),
-  sub(Depth, 1, RestDepth),
-  plan(Next, Goal, RestDepth, [Next|Visited], Moves, Final).
-
-five_move_plan(Moves, Final) :-
-  initial(Start),
-  goal(Goal),
-  sort(Start, SortedStart),
-  sort(Goal, SortedGoal),
-  plan(SortedStart, SortedGoal, 5, [SortedStart], Moves, Final).
-
-status(blocks_world, planned) :-
-  once(five_move_plan(_Moves, _Final)).
-
-plan(blocks_world, Moves) :-
-  once(five_move_plan(Moves, _Final)).
-
-finalState(blocks_world, Final) :-
-  once(five_move_plan(_Moves, Final)).
-
-blockCount(blocks_world, Count) :-
-  findall(Block, block(Block), Blocks),
-  length(Blocks, Count).

package/examples/bmi.pl

@@ -1,232 +0,0 @@
-% BMI — ARC-style Body Mass Index example adapted from Eyeling.
-%
-% The example normalizes metric or US inputs, computes BMI, assigns the WHO
-% adult category, derives a healthy-weight band for the same height, and emits
-% an inspectable report plus independent checks.
-%
-% For reproducibility and documentation only; not medical advice.
-
-% Editable metric input.
-% Output declarations: materialize/2 selects the relations written to this example's golden output.
-materialize(unitSystem, 2).
-materialize(weight, 2).
-materialize(height, 2).
-materialize(weightKg, 2).
-materialize(heightM, 2).
-materialize(units, 2).
-materialize(heightSquared, 2).
-materialize(bmi, 2).
-materialize(bmiRoundedInt, 2).
-materialize(healthyMinKg, 2).
-materialize(healthyMaxKg, 2).
-materialize(healthyMinKgRoundedInt, 2).
-materialize(healthyMaxKgRoundedInt, 2).
-materialize(category, 2).
-materialize(heightCm, 2).
-materialize(formula, 2).
-materialize(calculation, 2).
-materialize(categoryRule, 2).
-materialize(unitsExplanation, 2).
-materialize(c1, 2).
-materialize(c2, 2).
-materialize(c3, 2).
-materialize(c4, 2).
-materialize(c5, 2).
-materialize(c6, 2).
-materialize(c7, 2).
-materialize(c8, 2).
-materialize(c9, 2).
-materialize(result, 2).
-materialize(healthyWeightRangeKg, 2).
-materialize(checkPassed, 2).
-
-% Program structure: facts set up the scenario, and rules derive the materialized conclusions.
-unitSystem(input, metric).
-weight(input, 72.0).
-height(input, 178.0).
-
-% US alternative:
-% unitSystem(input, us).
-% weight(input, 158.73).
-% height(input, 70.08).
-
-% Normalization and BMI calculation.
-% Derivation rules: each rule below contributes one logical step toward the displayed results.
-weightKg(case, W) :-
-  unitSystem(input, metric),
-  weight(input, W).
-
-heightM(case, M) :-
-  unitSystem(input, metric),
-  height(input, H),
-  div(H, 100.0, M).
-
-units(reason, "Inputs were already metric, so kilograms stay kilograms and centimeters are divided by 100 to obtain meters.") :-
-  unitSystem(input, metric).
-
-weightKg(case, Kg) :-
-  unitSystem(input, us),
-  weight(input, W),
-  mul(W, 0.45359237, Kg).
-
-heightM(case, M) :-
-  unitSystem(input, us),
-  height(input, H),
-  mul(H, 0.0254, M).
-
-units(reason, "US inputs were converted to SI units: pounds to kilograms and inches to meters.") :-
-  unitSystem(input, us).
-
-heightSquared(case, M2) :-
-  heightM(case, M),
-  mul(M, M, M2).
-
-bmi(case, Bmi) :-
-  weightKg(case, Kg),
-  heightSquared(case, M2),
-  div(Kg, M2, Bmi).
-
-bmiRoundedInt(case, BmiRoundedInt) :-
-  bmi(case, Bmi),
-  mul(Bmi, 100.0, BmiX100),
-  rounded(BmiX100, BmiRoundedInt).
-
-healthyMinKg(case, HealthyMin) :-
-  heightSquared(case, M2),
-  mul(18.5, M2, HealthyMin).
-
-healthyMaxKg(case, HealthyMax) :-
-  heightSquared(case, M2),
-  mul(24.9, M2, HealthyMax).
-
-healthyMinKgRoundedInt(case, MinRoundedInt) :-
-  healthyMinKg(case, HealthyMin),
-  mul(HealthyMin, 10.0, MinX10),
-  rounded(MinX10, MinRoundedInt).
-
-healthyMaxKgRoundedInt(case, MaxRoundedInt) :-
-  healthyMaxKg(case, HealthyMax),
-  mul(HealthyMax, 10.0, MaxX10),
-  rounded(MaxX10, MaxRoundedInt).
-
-% WHO adult categories, using half-open intervals.
-category(decision, "Underweight") :-
-  bmi(case, Bmi),
-  lt(Bmi, 18.5).
-
-category(decision, "Normal") :-
-  bmi(case, Bmi),
-  ge(Bmi, 18.5),
-  lt(Bmi, 25.0).
-
-category(decision, "Overweight") :-
-  bmi(case, Bmi),
-  ge(Bmi, 25.0),
-  lt(Bmi, 30.0).
-
-category(decision, "Obesity I") :-
-  bmi(case, Bmi),
-  ge(Bmi, 30.0),
-  lt(Bmi, 35.0).
-
-category(decision, "Obesity II") :-
-  bmi(case, Bmi),
-  ge(Bmi, 35.0),
-  lt(Bmi, 40.0).
-
-category(decision, "Obesity III") :-
-  bmi(case, Bmi),
-  ge(Bmi, 40.0).
-
-% Answer and reason why.
-bmi(answer, 22.72) :-
-  bmiRoundedInt(case, 2272).
-
-category(answer, Category) :-
-  category(decision, Category).
-
-healthyMinKg(answer, 58.6) :-
-  healthyMinKgRoundedInt(case, 586).
-
-healthyMaxKg(answer, 78.9) :-
-  healthyMaxKgRoundedInt(case, 789).
-
-heightCm(answer, CmRounded) :-
-  heightM(case, M),
-  mul(M, 100.0, Cm),
-  rounded(Cm, CmRounded).
-
-formula(reason, "BMI is defined as weight in kilograms divided by height in meters squared.") :-
-  bmi(case, _Bmi).
-
-calculation(reason, "The normalized weight and height were used to compute BMI, then the result was mapped to the WHO adult category table.") :-
-  category(decision, _Category).
-
-categoryRule(reason, Category) :-
-  category(decision, Category).
-
-unitsExplanation(reason, Units) :-
-  units(reason, Units).
-
-% Independent checks.
-c1(check, "OK - the input was normalized into positive SI values.") :-
-  weightKg(case, Kg),
-  heightM(case, M),
-  gt(Kg, 0),
-  gt(M, 0).
-
-c2(check, "OK - height squared was reconstructed from the normalized height.") :-
-  heightM(case, M),
-  heightSquared(case, M2),
-  mul(M, M, M2).
-
-c3(check, "OK - the BMI value matches the BMI = kg / m² formula.") :-
-  weightKg(case, Kg),
-  heightSquared(case, M2),
-  bmi(case, Bmi),
-  div(Kg, M2, Bmi).
-
-c4(check, "OK - a BMI of 18.49 stays below the normal-weight threshold.") :-
-  lt(18.49, 18.5).
-
-c5(check, "OK - the lower boundary is half-open: BMI 18.5 is classified as Normal.") :-
-  ge(18.5, 18.5),
-  lt(18.5, 25.0).
-
-c6(check, "OK - BMI 25.0 starts the Overweight category.") :-
-  ge(25.0, 25.0),
-  lt(25.0, 30.0).
-
-c7(check, "OK - BMI 30.0 starts the Obesity I category.") :-
-  ge(30.0, 30.0),
-  lt(30.0, 35.0).
-
-c8(check, "OK - classification behavior is monotonic across representative BMI values.") :-
-  ge(22.0, 18.5),
-  lt(22.0, 25.0),
-  ge(27.0, 25.0),
-  lt(27.0, 30.0),
-  ge(41.0, 40.0).
-
-c9(check, "OK - the healthy-weight band was reconstructed from BMI 18.5 to 24.9 at the same height.") :-
-  heightSquared(case, M2),
-  healthyMinKg(case, Min),
-  healthyMaxKg(case, Max),
-  mul(18.5, M2, Min),
-  mul(24.9, M2, Max).
-
-% Derived report summary.  These relations are consequences of the calculation
-% and checks, not pre-written report lines.
-result(report, bmi(Bmi, Category)) :-
-  bmi(answer, Bmi),
-  category(answer, Category).
-
-healthyWeightRangeKg(report, range(Min, Max)) :-
-  healthyMinKg(answer, Min),
-  healthyMaxKg(answer, Max).
-
-heightCm(report, Height) :-
-  heightCm(answer, Height).
-
-checkPassed(report, Check) :-
-  statement(check, Check, _Message).