import numpy as np, json, urllib.request
SP="/private/tmp/claude-501/-Users-timlingo/82369039-a20e-4b5a-8a5e-28234a57b996/scratchpad"
np.seterr(all='ignore')
M=np.load(SP+'/emb.npy'); eids=open(SP+'/ids.txt',encoding='utf-8',errors='surrogateescape').read().split('\n')
eidx={k:i for i,k in enumerate(eids)}
gold=json.load(open("/Users/timlingo/Development/neuron-technologies/_wt-assoc-leg/tools/retrieval-eval/gold_set.json"))['queries']
VALS=sorted({r for q in gold if q['category']=='paraphrase' for r in q['relevant']})
VI=[eidx[v] for v in VALS]
def emb(t):
    b=json.dumps({"model":"nomic-embed-text","prompt":t}).encode()
    r=urllib.request.Request("http://127.0.0.1:11434/api/embeddings",data=b,headers={"Content-Type":"application/json"})
    v=np.array(json.load(urllib.request.urlopen(r,timeout=60))["embedding"],dtype=np.float32)
    return v/(np.linalg.norm(v)+1e-9)
print("qid  cat         bestValueNodeGlobalRank   goldGlobalRank  goldSiblingRank")
for q in gold:
    if q['category']!='paraphrase': continue
    v=emb(q['query']); s=M@v; s[~np.isfinite(s)]=-1
    ranks=sorted(int((s>s[j]).sum())+1 for j in VI)
    g=eidx[q['relevant'][0]]; gr=int((s>s[g]).sum())+1
    sv=np.array([s[j] for j in VI]); sib=int((sv>s[g]).sum())+1
    print("%-4s %-11s best=%-5d (top3 val ranks %s)  gold=%-5d sib=%d" % (q['id'],q['category'],ranks[0],ranks[:3],gr,sib))