"""Declared finite analytical controls; no neural training or source benchmark reproduction."""
import json,math,itertools,hashlib
from pathlib import Path
ROOT=Path(__file__).resolve().parents[1]
checks=[]
cases=0
def check(ok,label):
 global cases
 assert ok,label
 cases+=1
 if label not in checks: checks.append(label)
def fig(fid,title,xlabel,ylabel,x,series,caption):
 return dict(id=fid,title=title,kind='analytical_curve',x_label=xlabel,y_label=ylabel,x={'name':xlabel,'values':x},series=[{'name':n,'values':v}for n,v in series],caption=caption)
def pars(g,varc,mu=1.):
 tau=g/varc+1-g
 return (tau,None,None) if tau<=0 else (tau,g*mu/(varc*tau),1/tau)
for vc in [.25,.5,1.,4.]:
 for g in [0,.25,.5,1,1.2,1.4,2,4]:
  tau,mu,var=pars(g,vc)
  if tau<=0:
   check(mu is None and var is None,'nonpositive precision refuses density')
   continue
  for x in [-3,-.2,0,1,3]:
   guided=-x+g*(-(x-1)/vc+x)
   completed=-(x-mu)/var
   check(abs(guided-completed)<1e-11,'Gaussian guided score equals completed-square score')
  if g==0:check(abs(mu)<1e-12 and abs(var-1)<1e-12,'g0 recovers unconditional')
  if g==1:check(abs(mu-1)<1e-12 and abs(var-vc)<1e-12,'g1 recovers conditional')
check(abs(pars(4/3,4)[0])<1e-12,'wider conditional precision boundary')
gs=[0,.25,.5,.75,1,1.25,1.5,2,3,4];valid=[0,.25,.5,.75,1,1.1,1.2,1.25,1.3]
figures=[fig('guided-gaussian','Guidance changes mean and spread','Guidance g (g=1 conditional)','Mean / variance',gs,[('Mean',[pars(g,.25)[1]for g in gs]),('Variance',[pars(g,.25)[2]for g in gs])],'Exact fixed-noise p_u=N(0,1), p_c=N(1,0.25), score s_u+g(s_c-s_u). Not an endpoint law for a complete diffusion sampler.'),
fig('normalizability-boundary','A wider conditional can lose normalizability','Guidance g','Precision tau',gs,[('Conditional variance4',[pars(g,4)[0]for g in gs]),('Zero precision boundary',[0]*len(gs))],'Exact fixed-noise construction with p_c=N(1,4). The power product is normalizable only where tau>0; g>=4/3 is invalid. No finite variance is assigned to invalid points.')]
out={'article_slug':'diffusion-guidance-bias','figures':figures}
(ROOT/'figures').mkdir(exist_ok=True)
(ROOT/'figures/diffusion-guidance-bias.json').write_text(json.dumps(out,indent=2)+'\n',encoding='utf-8')
receipt={'status':'PASS','checks':checks,'cases':cases,'empirical_benchmark':False,'source_proofs_verified':False,'figure_sha256':hashlib.sha256((ROOT/'figures/diffusion-guidance-bias.json').read_bytes()).hexdigest()}
(ROOT/'analysis/diffusion-guidance-bias.receipt.json').write_text(json.dumps(receipt,indent=2)+'\n',encoding='utf-8')
print(json.dumps(receipt))