dexbot 1.1.10

package/dist/analysis/ama_fitting/optimizer_high_resolution.js

@@ -0,0 +1,782 @@
+'use strict';
+Object.defineProperty(exports, "__esModule", { value: true });
+const fs = require('fs');
+const path = require('path');
+const os = require('os');
+const { Worker, isMainThread, parentPort, workerData } = require('worker_threads');
+const { calculateAMA } = require('../../market_adapter/core/strategies/ama');
+const { toIntervalLabel } = require('../../market_adapter/interval_utils');
+const { generateHTML } = require('../../market_adapter/lp_chart_core');
+const { loadLpDataFile, } = require('../../market_adapter/lp_chart_runner');
+const { PATHS } = require('../../modules/paths');
+const AMA_PROFILES_FILE = PATHS.PROFILES.MARKET_PROFILES_JSON;
+/**
+ * AMA GEOMETRIC OPTIMIZER
+ *
+ * Finds AMA parameters (ER, Fast, Slow) using geometric metrics only.
+ *
+ * Objective flow:
+ *   1. Compute a per-AMA max-distance cap from the max candle deviation percentile.
+ *   2. Keep only candidates at or below that cap.
+ *   3. Select the candidate with the lowest additive linear score.
+ *
+ * Distance to price is still tracked for reporting.
+ * Winner selection uses an additive tradeoff between AMA movement and
+ * AMA-to-price distance, so candidates must stay close enough while also
+ * being smooth.
+ *
+ * Usage (build first, then run compiled JS):
+ *   npm run build && node dist/analysis/ama_fitting/optimizer_high_resolution.js --data <path-to-lp-candles.json>
+ *   node dist/analysis/ama_fitting/optimizer_high_resolution.js --data <path-to-lp-candles.json> --write-profiles
+ *   node dist/analysis/ama_fitting/optimizer_high_resolution.js --data <path-to-lp-candles.json> --erMax 400 --fastMax 20 --slowMax 200
+ *
+ */
+const DEFAULT_SEARCH = {
+    er: { min: 500, max: 1000, count: 15, step: null, quantum: 1 },
+    fast: { min: 2, max: 8, count: 30, step: null, quantum: 0.01 },
+    slow: { min: 35, max: 140, count: 30, step: null, quantum: 0.1 },
+};
+// ── Default per-AMA distance weights ────────────────────────────────────────
+// Each AMA has its own λ that controls the movement-vs-distance tradeoff.
+// Override individually via --ama1Weight, --ama2Weight, etc. CLI flags.
+const AMA_OBJECTIVES = [
+    { key: 'AMA1', name: 'AMA1 (min move, cap 25%)', distanceCapQuantile: 0.25, distanceWeight: 0.0031 },
+    { key: 'AMA2', name: 'AMA2 (min move, cap 30%)', distanceCapQuantile: 0.30, distanceWeight: 0.0025 },
+    { key: 'AMA3', name: 'AMA3 (min move, cap 35%)', distanceCapQuantile: 0.35, distanceWeight: 0.00185 },
+    { key: 'AMA4', name: 'AMA4 (min move, cap 40%)', distanceCapQuantile: 0.40, distanceWeight: 0.0013 },
+];
+function cloneObjectives() {
+    return AMA_OBJECTIVES.map((o) => ({ ...o }));
+}
+// ── Parameter ranges ──────────────────────────────────────────────────────────
+function quantize(value, quantum) {
+    if (!Number.isFinite(quantum) || quantum <= 0)
+        return value;
+    return Math.round(value / quantum) * quantum;
+}
+function geometricRange(min, max, count, quantum = null) {
+    const out = [];
+    const ratio = Math.pow(max / min, 1 / (count - 1));
+    for (let i = 0; i < count; i++) {
+        let v = min * Math.pow(ratio, i);
+        if (i === 0)
+            v = min;
+        if (i === count - 1)
+            v = max;
+        v = quantize(v, quantum);
+        v = Math.max(min, Math.min(max, v));
+        out.push(parseFloat(v.toFixed(6)));
+    }
+    return [...new Set(out)].sort((a, b) => a - b);
+}
+function buildDimension(label, cfg) {
+    const min = Number(cfg.min);
+    const max = Number(cfg.max);
+    const step = Number(cfg.step);
+    const count = Number(cfg.count);
+    const quantum = Number(cfg.quantum);
+    if (Number.isFinite(step) && step > 0) {
+        const values = range(min, max, step, 2);
+        return {
+            values,
+            meta: {
+                mode: 'linear',
+                min: values[0],
+                max: values[values.length - 1],
+                step,
+                quantum: Number.isFinite(quantum) && quantum > 0 ? quantum : null,
+                count: values.length,
+                ratio: null,
+            },
+        };
+    }
+    const ratio = Math.pow(max / min, 1 / (count - 1));
+    const values = geometricRange(min, max, count, quantum);
+    return {
+        values,
+        meta: {
+            mode: 'geometric',
+            min: values[0],
+            max: values[values.length - 1],
+            step: null,
+            quantum: Number.isFinite(quantum) && quantum > 0 ? quantum : null,
+            count: values.length,
+            ratio,
+        },
+    };
+}
+function parseArgs(argv = process.argv.slice(2)) {
+    const args = Array.isArray(argv) ? argv : [];
+    const out = {
+        dataFile: null,
+        er: { ...DEFAULT_SEARCH.er },
+        fast: { ...DEFAULT_SEARCH.fast },
+        slow: { ...DEFAULT_SEARCH.slow },
+        workers: null,
+        writeProfiles: false,
+        fixedEr: null,
+        fixedFast: null,
+        ama1Cap: null,
+        ama2Cap: null,
+        ama3Cap: null,
+        ama4Cap: null,
+        ama1Weight: null,
+        ama2Weight: null,
+        ama3Weight: null,
+        ama4Weight: null,
+    };
+    for (let i = 0; i < args.length; i++) {
+        const a = args[i];
+        const v = args[i + 1];
+        switch (a) {
+            case '--data':
+                out.dataFile = v || null;
+                i++;
+                break;
+            case '--fixEr': {
+                const n = Number(v);
+                if (!Number.isFinite(n))
+                    throw new Error(`--fixEr requires a finite number, got "${v}"`);
+                out.fixedEr = n;
+                i++;
+                break;
+            }
+            case '--fixFast': {
+                const n = Number(v);
+                if (!Number.isFinite(n))
+                    throw new Error(`--fixFast requires a finite number, got "${v}"`);
+                out.fixedFast = n;
+                i++;
+                break;
+            }
+            case '--erMin':
+                out.er.min = Number(v);
+                i++;
+                break;
+            case '--erMax':
+                out.er.max = Number(v);
+                i++;
+                break;
+            case '--erStep':
+                out.er.step = Number(v);
+                i++;
+                break;
+            case '--erCount':
+                out.er.count = Number(v);
+                i++;
+                break;
+            case '--fastMin':
+                out.fast.min = Number(v);
+                i++;
+                break;
+            case '--fastMax':
+                out.fast.max = Number(v);
+                i++;
+                break;
+            case '--fastStep':
+                out.fast.step = Number(v);
+                i++;
+                break;
+            case '--fastCount':
+                out.fast.count = Number(v);
+                i++;
+                break;
+            case '--slowMin':
+                out.slow.min = Number(v);
+                i++;
+                break;
+            case '--slowMax':
+                out.slow.max = Number(v);
+                i++;
+                break;
+            case '--slowStep':
+                out.slow.step = Number(v);
+                i++;
+                break;
+            case '--slowCount':
+                out.slow.count = Number(v);
+                i++;
+                break;
+            case '--ama1Cap':
+                out.ama1Cap = Number(v);
+                i++;
+                break;
+            case '--ama2Cap':
+                out.ama2Cap = Number(v);
+                i++;
+                break;
+            case '--ama3Cap':
+                out.ama3Cap = Number(v);
+                i++;
+                break;
+            case '--ama4Cap':
+                out.ama4Cap = Number(v);
+                i++;
+                break;
+            case '--ama1Weight':
+                out.ama1Weight = Number(v);
+                i++;
+                break;
+            case '--ama2Weight':
+                out.ama2Weight = Number(v);
+                i++;
+                break;
+            case '--ama3Weight':
+                out.ama3Weight = Number(v);
+                i++;
+                break;
+            case '--ama4Weight':
+                out.ama4Weight = Number(v);
+                i++;
+                break;
+            case '--workers':
+                out.workers = Number(v);
+                i++;
+                break;
+            case '--write-profiles':
+                out.writeProfiles = true;
+                break;
+        }
+    }
+    return out;
+}
+function percentile(values, q) {
+    if (!Array.isArray(values) || values.length === 0)
+        return null;
+    const sorted = [...values].sort((a, b) => a - b);
+    const qq = Math.max(0, Math.min(1, q));
+    const pos = (sorted.length - 1) * qq;
+    const lo = Math.floor(pos);
+    const hi = Math.ceil(pos);
+    if (lo === hi)
+        return sorted[lo];
+    const t = pos - lo;
+    return sorted[lo] * (1 - t) + sorted[hi] * t;
+}
+function getAmaObjectivesFromArgs(args) {
+    const objectives = cloneObjectives();
+    for (const o of objectives) {
+        const id = o.key.toLowerCase();
+        const cap = args[`${id}Cap`];
+        if (Number.isFinite(cap))
+            o.distanceCapQuantile = cap;
+        if (!Number.isFinite(o.distanceCapQuantile) || o.distanceCapQuantile <= 0 || o.distanceCapQuantile > 1) {
+            throw new Error(`Invalid cap for ${o.key}: ${o.distanceCapQuantile}. Use 0 < cap <= 1`);
+        }
+        const w = args[`${id}Weight`];
+        if (Number.isFinite(w)) {
+            if (w <= 0)
+                throw new Error(`${o.key} distanceWeight must be positive, got ${w}`);
+            o.distanceWeight = w;
+        }
+    }
+    return objectives;
+}
+function ensureValidRange(label, cfg) {
+    const { min, max, step, count } = cfg;
+    if (!Number.isFinite(min) || !Number.isFinite(max) || max < min) {
+        throw new Error(`Invalid ${label} range: min=${min}, max=${max}`);
+    }
+    if (Number.isFinite(step) && step > 0)
+        return;
+    if (!Number.isFinite(count) || count < 2) {
+        throw new Error(`Invalid ${label} geometric sampling count: ${count}`);
+    }
+}
+function boundaryFlags(winner, erValues, fastValues, slowValues, fixedEr = null, fixedFast = null) {
+    if (!winner)
+        return { er: null, fast: null, slow: null, any: false };
+    const eps = 1e-9;
+    const minEr = erValues[0], maxEr = erValues[erValues.length - 1];
+    const minFast = fastValues[0], maxFast = fastValues[fastValues.length - 1];
+    const minSlow = slowValues[0], maxSlow = slowValues[slowValues.length - 1];
+    const er = Number.isFinite(fixedEr) ? null : (Math.abs(winner.er - minEr) < eps ? 'min' : (Math.abs(winner.er - maxEr) < eps ? 'max' : null));
+    const fast = Number.isFinite(fixedFast) ? null : (Math.abs(winner.fast - minFast) < eps ? 'min' : (Math.abs(winner.fast - maxFast) < eps ? 'max' : null));
+    const slow = Math.abs(winner.slow - minSlow) < eps ? 'min' : (Math.abs(winner.slow - maxSlow) < eps ? 'max' : null);
+    return { er, fast, slow, any: !!(er || fast || slow) };
+}
+const { ensureDir } = require('../../modules/order/utils/system');
+const { range } = require('../math_utils');
+const { readJSON, writeJSON } = require('../../modules/utils/fs_utils');
+// ── Data loaders ──────────────────────────────────────────────────────────────
+function normalizeSymbol(value) {
+    return String(value || '').trim().toUpperCase();
+}
+function inferIntervalLabel(dataFile, meta) {
+    const fromMeta = Number(meta?.intervalSeconds);
+    if (Number.isFinite(fromMeta) && fromMeta > 0) {
+        return toIntervalLabel(fromMeta);
+    }
+    const m = String(path.basename(dataFile || '')).match(/_(\d+)([mhd])\.json$/i);
+    return m ? `${m[1]}${m[2].toLowerCase()}` : '1h';
+}
+function updateAmaProfilesFile({ dataFile, meta, winners, sourceResultsFile }) {
+    const assetASymbol = normalizeSymbol(meta?.assetA?.symbol);
+    const assetBSymbol = normalizeSymbol(meta?.assetB?.symbol);
+    const assetAId = normalizeSymbol(meta?.assetA?.id);
+    const assetBId = normalizeSymbol(meta?.assetB?.id);
+    const assetA = assetASymbol || assetAId;
+    const assetB = assetBSymbol || assetBId;
+    if (!assetA || !assetB)
+        return;
+    const intervalSeconds = Number(meta?.intervalSeconds);
+    const intervalLabel = inferIntervalLabel(dataFile, meta);
+    const key = `${assetA}|${assetB}|${Number.isFinite(intervalSeconds) && intervalSeconds > 0 ? intervalSeconds : intervalLabel}`;
+    const payload = {
+        version: 1,
+        updatedAt: new Date().toISOString(),
+        profiles: [],
+    };
+    if (fs.existsSync(AMA_PROFILES_FILE)) {
+        try {
+            const current = readJSON(AMA_PROFILES_FILE);
+            if (current && typeof current === 'object') {
+                payload.version = Number(current.version) || 1;
+                payload.profiles = Array.isArray(current.profiles) ? current.profiles : [];
+            }
+        }
+        catch (_) { }
+    }
+    const profile = {
+        key,
+        assetA,
+        assetB,
+        assetAId: assetAId || null,
+        assetBId: assetBId || null,
+        poolId: meta?.pool || null,
+        intervalSeconds: Number.isFinite(intervalSeconds) && intervalSeconds > 0 ? intervalSeconds : null,
+        intervalLabel,
+        defaultAma: 'AMA3',
+        sourceResultsFile,
+        updatedAt: payload.updatedAt,
+        amas: {
+            AMA1: {
+                name: winners.ama1.label,
+                erPeriod: winners.ama1.er,
+                fastPeriod: winners.ama1.fast,
+                slowPeriod: winners.ama1.slow,
+            },
+            AMA2: {
+                name: winners.ama2.label,
+                erPeriod: winners.ama2.er,
+                fastPeriod: winners.ama2.fast,
+                slowPeriod: winners.ama2.slow,
+            },
+            AMA3: {
+                name: winners.ama3.label,
+                erPeriod: winners.ama3.er,
+                fastPeriod: winners.ama3.fast,
+                slowPeriod: winners.ama3.slow,
+            },
+            AMA4: {
+                name: winners.ama4.label,
+                erPeriod: winners.ama4.er,
+                fastPeriod: winners.ama4.fast,
+                slowPeriod: winners.ama4.slow,
+            },
+        },
+    };
+    const idx = payload.profiles.findIndex((p) => String(p?.key || '') === key);
+    if (idx >= 0)
+        payload.profiles[idx] = profile;
+    else
+        payload.profiles.push(profile);
+    ensureDir(path.dirname(AMA_PROFILES_FILE));
+    writeJSON(AMA_PROFILES_FILE, payload);
+}
+function calcTotalAmaMovement(amaValues, erPeriod) {
+    const skip = erPeriod + 1;
+    let total = 0;
+    for (let i = skip + 1; i < amaValues.length; i++) {
+        total += Math.abs(amaValues[i] - amaValues[i - 1]) / amaValues[i - 1];
+    }
+    return total;
+}
+// ── Informational: area above/below AMA ──────────────────────────────────────
+function calcArea(amaValues, candles, erPeriod) {
+    const skip = erPeriod + 1;
+    let above = 0, below = 0, maxUp = 0, maxDown = 0;
+    for (let i = skip; i < candles.length; i++) {
+        const ama = amaValues[i];
+        if (candles[i].high > ama) {
+            const d = (candles[i].high - ama) / ama;
+            above += d;
+            if (d > maxUp)
+                maxUp = d;
+        }
+        if (candles[i].low < ama) {
+            const d = (ama - candles[i].low) / ama;
+            below += d;
+            if (d > maxDown)
+                maxDown = d;
+        }
+    }
+    const total = above + below;
+    const maxDist = Math.max(maxUp, maxDown);
+    return { above, below, total, maxUp, maxDown, maxDist };
+}
+function calcTotalRelativeDistance(amaValues, candles, erPeriod) {
+    const skip = erPeriod + 1;
+    let total = 0;
+    for (let i = skip; i < candles.length; i++) {
+        const ama = amaValues[i];
+        const close = candles[i].close;
+        total += Math.abs(close - ama) / ama;
+    }
+    return total;
+}
+function runSearchShard(payload, onProgress = null) {
+    const { workerId, erShard, fastValues, slowValues, candles, closes, } = payload;
+    const totalCombos = erShard.length * fastValues.length * slowValues.length;
+    const progressStep = Math.max(2000, Math.floor(totalCombos / 20));
+    let checked = 0;
+    let valid = 0;
+    const entries = [];
+    const startMs = Date.now();
+    for (const er of erShard) {
+        for (const fast of fastValues) {
+            for (const slow of slowValues) {
+                checked++;
+                if (fast >= slow) {
+                    if (onProgress && (checked % progressStep === 0 || checked === totalCombos)) {
+                        const elapsedSec = (Date.now() - startMs) / 1000;
+                        onProgress({ workerId, checked, total: totalCombos, elapsedSec });
+                    }
+                    continue;
+                }
+                valid++;
+                const ama = calculateAMA(closes, { erPeriod: er, fastPeriod: fast, slowPeriod: slow });
+                const area = calcArea(ama, candles, er);
+                const amaMovementTotal = calcTotalAmaMovement(ama, er);
+                const distanceTotal = calcTotalRelativeDistance(ama, candles, er);
+                const bandFactorPct = area.maxDist * 200;
+                const entry = {
+                    er, fast, slow,
+                    area,
+                    amaMovementTotal,
+                    bandFactorPct,
+                    distanceTotal,
+                };
+                entries.push(entry);
+                if (onProgress && (checked % progressStep === 0 || checked === totalCombos)) {
+                    const elapsedSec = (Date.now() - startMs) / 1000;
+                    onProgress({ workerId, checked, total: totalCombos, elapsedSec });
+                }
+            }
+        }
+    }
+    return { workerId, entries, totalCombos, validCombos: valid };
+}
+function spawnShardWorker(payload, onProgress) {
+    return new Promise((resolve, reject) => {
+        const worker = new Worker(__filename, { workerData: { type: 'search_shard', payload } });
+        worker.on('message', (msg) => {
+            if (!msg || typeof msg !== 'object')
+                return;
+            if (msg.type === 'progress') {
+                if (onProgress)
+                    onProgress(msg);
+                return;
+            }
+            if (msg.type === 'done')
+                resolve(msg.result);
+        });
+        worker.on('error', reject);
+        worker.on('exit', (code) => {
+            if (code !== 0)
+                reject(new Error(`Worker exited with code ${code}`));
+        });
+    });
+}
+function splitIntoShards(values, shardCount) {
+    const out = [];
+    const size = Math.ceil(values.length / shardCount);
+    for (let i = 0; i < values.length; i += size) {
+        out.push(values.slice(i, i + size));
+    }
+    return out.filter((s) => s.length > 0);
+}
+// ── Main ───────────────────────────────────────────────────────────────────────
+async function run() {
+    const args = parseArgs();
+    let dataFile = args.dataFile;
+    const objectives = getAmaObjectivesFromArgs(args);
+    ensureValidRange('ER', args.er);
+    ensureValidRange('Fast', args.fast);
+    ensureValidRange('Slow', args.slow);
+    const erDim = buildDimension('ER', args.er);
+    const fastDim = buildDimension('Fast', args.fast);
+    const slowDim = buildDimension('Slow', args.slow);
+    const ER_VALUES = Number.isFinite(args.fixedEr) ? [args.fixedEr] : erDim.values;
+    const FAST_VALUES = Number.isFinite(args.fixedFast) ? [args.fixedFast] : fastDim.values;
+    const SLOW_VALUES_AREA = slowDim.values;
+    const totalCombos = ER_VALUES.length * FAST_VALUES.length * SLOW_VALUES_AREA.length;
+    const startMs = Date.now();
+    const erLabel = Number.isFinite(args.fixedEr) ? `fixed(${args.fixedEr})` : `${ER_VALUES[0]}–${ER_VALUES[ER_VALUES.length - 1]}`;
+    const fastLabel = Number.isFinite(args.fixedFast) ? `fixed(${args.fixedFast})` : `${FAST_VALUES[0]}–${FAST_VALUES[FAST_VALUES.length - 1]}`;
+    console.log('================================================================================');
+    console.log(' AMA GEOMETRIC OPTIMIZER');
+    console.log('================================================================================');
+    console.log(`  4 AMAs — pure geometric, no grid or bot settings`);
+    console.log('  Objective: minimise movement + λ·distance under per-AMA max distance caps');
+    for (const o of objectives) {
+        console.log(`  ${o.key}: distance cap q=${o.distanceCapQuantile.toFixed(2)}  λ=${o.distanceWeight.toFixed(6)}`);
+    }
+    console.log(`  Ranges:     ER ${erLabel}  Fast ${fastLabel}  Slow ${SLOW_VALUES_AREA[0]}–${SLOW_VALUES_AREA[SLOW_VALUES_AREA.length - 1]}`);
+    if (!Number.isFinite(args.fixedEr) && !Number.isFinite(args.fixedFast)) {
+        console.log(`  Sampling:   ER ${erDim.meta.mode}(${erDim.meta.count}${erDim.meta.ratio ? `,x${erDim.meta.ratio.toFixed(3)}` : ''})  Fast ${fastDim.meta.mode}(${fastDim.meta.count}${fastDim.meta.ratio ? `,x${fastDim.meta.ratio.toFixed(3)}` : ''})  Slow ${slowDim.meta.mode}(${slowDim.meta.count}${slowDim.meta.ratio ? `,x${slowDim.meta.ratio.toFixed(3)}` : ''})`);
+    }
+    console.log(`  Combos:     ${totalCombos}\n`);
+    // Load data
+    let candles, dataLabel, dataMeta = null;
+    if (!dataFile) {
+        throw new Error('Optimizer requires --data <lp_pool_*.json>. Use --write-profiles to also update profiles/market_profiles.json.');
+    }
+    const loaded = loadLpDataFile(path.resolve(dataFile));
+    candles = loaded.candleObjects;
+    const m = loaded.meta;
+    dataMeta = m;
+    dataLabel = m ? `LP Pool ${m.pool} (${m.assetA?.symbol}/${m.assetB?.symbol})` : path.basename(dataFile);
+    const closes = candles.map(c => c.close);
+    console.log(`  Data:       ${dataLabel}  (${candles.length} candles)\n`);
+    // ── Run: core-parallel full-grid scan split by ER shards ──────────────────
+    const cpuCount = Math.max(1, os.cpus().length);
+    const requestedWorkers = Number.isFinite(args.workers) && args.workers > 0 ? Math.floor(args.workers) : cpuCount;
+    const workerCount = Math.max(1, Math.min(requestedWorkers, ER_VALUES.length));
+    const erShards = splitIntoShards(ER_VALUES, workerCount);
+    const shardTotals = erShards.map((shard) => shard.length * FAST_VALUES.length * SLOW_VALUES_AREA.length);
+    const shardProgress = new Map();
+    console.log(`  Parallel workers: ${erShards.length} shard workers (CPU cores available: ${cpuCount})`);
+    const shardResults = await Promise.all(erShards.map((erShard, idx) => {
+        const workerPayload = {
+            workerId: idx + 1,
+            erShard,
+            fastValues: FAST_VALUES,
+            slowValues: SLOW_VALUES_AREA,
+            candles,
+            closes,
+        };
+        return spawnShardWorker(workerPayload, (msg) => {
+            shardProgress.set(msg.workerId, msg.checked);
+            const done = shardTotals.reduce((acc, total, i) => acc + Math.min(total, shardProgress.get(i + 1) || 0), 0);
+            const pct = ((done / totalCombos) * 100).toFixed(1);
+            console.log(`  [scan] ${pct}%  (${done}/${totalCombos})  w${msg.workerId} ${msg.elapsedSec.toFixed(1)}s`);
+        });
+    }));
+    const entries = shardResults.flatMap((r) => r.entries);
+    const validCombos = shardResults.reduce((acc, r) => acc + r.validCombos, 0);
+    const maxDistVals = entries.map((e) => e.area.maxDist);
+    // When both ER and Fast are fixed, the search is 1D over Slow only. Slow
+    // variation rarely produces maxDist outliers, so the distance cap filter adds
+    // no value here — λ alone cleanly separates the candidates.
+    const skipCap = Number.isFinite(args.fixedEr) && Number.isFinite(args.fixedFast);
+    const objectiveResults = objectives.map((objective) => {
+        const distanceWeight = objective.distanceWeight;
+        const computedCap = skipCap ? null : percentile(maxDistVals, objective.distanceCapQuantile);
+        const cappedEntries = skipCap ? entries : entries.filter((e) => e.area.maxDist <= computedCap);
+        let best = null;
+        for (const e of cappedEntries) {
+            const score = e.amaMovementTotal + (distanceWeight * e.distanceTotal);
+            const bestScore = best ? best.weightedScore : Infinity;
+            if (!best || score < bestScore || (score === bestScore && e.amaMovementTotal < best.amaMovementTotal) || (score === bestScore && e.amaMovementTotal === best.amaMovementTotal && e.distanceTotal < best.distanceTotal)) {
+                best = {
+                    ...e,
+                    weightedScore: score,
+                    normDistance: null,
+                    normMovement: null,
+                    key: objective.key,
+                    label: objective.name,
+                };
+            }
+        }
+        return {
+            objective,
+            distanceWeight,
+            best,
+            totalCombos,
+            validCombos,
+            candidatesUnderCap: cappedEntries.length,
+            maxDistanceCap: computedCap,
+        };
+    });
+    const elapsedSec = (Date.now() - startMs) / 1000;
+    console.log(`  Completed parallel search in ${elapsedSec.toFixed(1)}s  (valid combos: ${validCombos})\n`);
+    const selected = objectiveResults.map((r) => r.best).filter(Boolean);
+    const failed = objectiveResults.filter((r) => !r.best).map((r) => r.objective?.key || '?');
+    if (failed.length > 0) {
+        throw new Error(`No candidate under distance cap for: ${failed.join(', ')}. Increase corresponding cap (e.g. --ama4Cap 0.35).`);
+    }
+    const ama1 = selected.find((s) => s.key === 'AMA1') || null;
+    const ama2 = selected.find((s) => s.key === 'AMA2') || null;
+    const ama3 = selected.find((s) => s.key === 'AMA3') || null;
+    const ama4 = selected.find((s) => s.key === 'AMA4') || null;
+    function detail(label, r, optimisedFor) {
+        if (!r) {
+            console.log(`  ${label}`);
+            console.log('  └─ No valid candidate under constraint\n');
+            return;
+        }
+        const asymmetry = Math.abs(r.area.above - r.area.below);
+        const bias = r.area.above > r.area.below ? 'AMA below price' : 'AMA above price';
+        console.log(`  ${label}`);
+        console.log(`  ├─ Optimised for:  ${optimisedFor}`);
+        console.log(`  ├─ Params:         ER=${r.er}  Fast=${r.fast}  Slow=${r.slow}`);
+        console.log(`  ├─ Area total:     ${r.area.total.toFixed(2)}  (above ${r.area.above.toFixed(2)}  below ${r.area.below.toFixed(2)})`);
+        console.log(`  ├─ Asymmetry:      ${asymmetry.toFixed(2)}  (${bias})`);
+        console.log(`  └─ Movement:       ${r.amaMovementTotal.toFixed(4)}\n`);
+    }
+    console.log('================================================================================');
+    console.log(` 4 AMAs  —  pure geometric  (${ER_VALUES.length}×${FAST_VALUES.length}×${SLOW_VALUES_AREA.length} combinations)`);
+    console.log('================================================================================\n');
+    for (const w of selected) {
+        detail(w.label, w, `score=${w.weightedScore.toFixed(6)}  λ=${objectiveResults.find((r) => r.objective?.key === w.key)?.distanceWeight?.toFixed(4) || 'n/a'}  rawDist=${w.distanceTotal.toFixed(2)}  rawMove=${w.amaMovementTotal.toFixed(2)}`);
+    }
+    for (const r of objectiveResults) {
+        if (Number.isFinite(r.maxDistanceCap)) {
+            console.log(`  ${r.objective.key} cap value: ${r.maxDistanceCap.toFixed(4)}  (candidates under cap: ${r.candidatesUnderCap})`);
+        }
+    }
+    console.log();
+    // ── Side-by-side summary ───────────────────────────────────────────────────
+    console.log('================================================================================');
+    console.log(' SUMMARY');
+    console.log('================================================================================\n');
+    console.log('                |  ER  | Fast | Slow | Dist    | Move    | Area    | MaxDist');
+    console.log('────────────────┼──────┼──────┼──────┼─────────┼─────────┼─────────┼─────────');
+    for (const r of selected) {
+        const name = r.key;
+        if (!r)
+            continue;
+        console.log(`${name.padEnd(15)} | ` +
+            `${r.er.toString().padStart(4)} | ` +
+            `${r.fast.toFixed(1).padStart(4)} | ` +
+            `${r.slow.toFixed(1).padStart(4)} | ` +
+            `${r.distanceTotal.toFixed(2).padStart(7)} | ` +
+            `${r.amaMovementTotal.toFixed(2).padStart(7)} | ` +
+            `${r.area.total.toFixed(2).padStart(7)} | ` +
+            `${(r.area.maxDist * 100).toFixed(1).padStart(7)}%`);
+    }
+    console.log();
+    const boundarySummary = {
+        AMA1: boundaryFlags(ama1, ER_VALUES, FAST_VALUES, SLOW_VALUES_AREA, args.fixedEr, args.fixedFast),
+        AMA2: boundaryFlags(ama2, ER_VALUES, FAST_VALUES, SLOW_VALUES_AREA, args.fixedEr, args.fixedFast),
+        AMA3: boundaryFlags(ama3, ER_VALUES, FAST_VALUES, SLOW_VALUES_AREA, args.fixedEr, args.fixedFast),
+        AMA4: boundaryFlags(ama4, ER_VALUES, FAST_VALUES, SLOW_VALUES_AREA, args.fixedEr, args.fixedFast),
+    };
+    console.log(' Boundary Check');
+    console.log('────────────────────────────────────────────────────────────────────────────────');
+    for (const [name, b] of Object.entries(boundarySummary)) {
+        const txt = b.any ? `ER:${b.er || '-'} Fast:${b.fast || '-'} Slow:${b.slow || '-'}` : 'none';
+        console.log(`${name.padEnd(24)} ${txt}`);
+    }
+    console.log();
+    // ── Chart ─────────────────────────────────────────────────────────────────
+    const ws = (v) => String(v).replace('.', '_');
+    const lambdaSuffix = `_w${objectives.map((o) => ws(String(o.distanceWeight))).join('_')}`;
+    const COLOR_CYCLE = ['#26a69a', '#fb8c00', '#5c9ee6', '#ef5350'];
+    const DASH_CYCLE = ['dot', 'solid', 'dash', 'dashdot'];
+    const candleArrays = candles.map(c => [c.timestamp, c.open, c.high, c.low, c.close, c.volume]);
+    const chartMeta = {
+        ...(dataMeta || {}),
+        assetA: dataMeta?.assetA || { symbol: '?' },
+        assetB: dataMeta?.assetB || { symbol: '?' },
+        intervalSeconds: dataMeta?.intervalSeconds || 3600,
+        fetchedAt: dataMeta?.fetchedAt || new Date().toISOString(),
+    };
+    const amaResults = selected.map((w, i) => {
+        const ama = calculateAMA(closes, { erPeriod: w.er, fastPeriod: w.fast, slowPeriod: w.slow });
+        return {
+            name: w.label,
+            erPeriod: w.er,
+            fastPeriod: w.fast,
+            slowPeriod: w.slow,
+            color: COLOR_CYCLE[i % COLOR_CYCLE.length],
+            dash: DASH_CYCLE[i % DASH_CYCLE.length],
+            lineWidth: i === 0 ? 2 : 1.5,
+            values: ama,
+        };
+    });
+    const chartName = dataFile
+        ? `optimization_chart_${path.basename(dataFile, '.json')}${lambdaSuffix}.html`
+        : `optimization_chart_high_resolution${lambdaSuffix}.html`;
+    const chartPath = path.join(PATHS.ANALYSIS.CHARTS_DIR, chartName);
+    ensureDir(path.dirname(chartPath));
+    const html = generateHTML(chartMeta, candleArrays, amaResults);
+    fs.writeFileSync(chartPath, html, 'utf8');
+    console.log(`  Chart:       ${path.relative(process.cwd(), chartPath)}\n`);
+    // ── Save ──────────────────────────────────────────────────────────────────
+    const outName = dataFile
+        ? `optimization_results_${path.basename(dataFile, '.json')}${lambdaSuffix}.json`
+        : `optimization_results_high_resolution${lambdaSuffix}.json`;
+    const outPath = path.join(__dirname, outName);
+    writeJSON(outPath, {
+        meta: {
+            dataLabel,
+            candles: candles.length,
+            totalCombos,
+            ranges: {
+                er: { ...erDim.meta },
+                fast: { ...fastDim.meta },
+                slow: { ...slowDim.meta },
+                fixedEr: Number.isFinite(args.fixedEr) ? args.fixedEr : null,
+                fixedFast: Number.isFinite(args.fixedFast) ? args.fixedFast : null,
+            },
+            boundaryFlags: boundarySummary,
+            objective: {
+                type: 'per_ama_distance_weights',
+                distanceMetric: 'sum(abs(close-ama)/ama)',
+                movementMetric: 'sum(abs(ama_t-ama_t-1)/ama_t-1)',
+                distanceWeightByAma: Object.fromEntries(objectiveResults.map((r) => [r.objective.key, r.distanceWeight])),
+                weights: objectives,
+                maxDistanceCap: {
+                    value: null,
+                    quantileByAma: Object.fromEntries(objectives.map((o) => [o.key, o.distanceCapQuantile])),
+                    appliedValueByAma: Object.fromEntries(objectiveResults.map((r) => [r.objective.key, r.maxDistanceCap])),
+                },
+            },
+            amas: {
+                AMA1: ama1,
+                AMA2: ama2,
+                AMA3: ama3,
+                AMA4: ama4,
+            },
+        },
+    });
+    if (args.writeProfiles && dataFile && ama1 && ama2 && ama3 && ama4) {
+        updateAmaProfilesFile({
+            dataFile,
+            meta: dataMeta,
+            winners: { ama1, ama2, ama3, ama4 },
+            sourceResultsFile: outName,
+        });
+    }
+    console.log(`================================================================================`);
+    console.log(`  Saved: ${outName}\n`);
+    if (args.writeProfiles && dataFile) {
+        console.log(`  Updated: ${path.relative(process.cwd(), AMA_PROFILES_FILE)}\n`);
+    }
+    else {
+        console.log(`  Profiles unchanged. Add --write-profiles to update ${path.relative(process.cwd(), AMA_PROFILES_FILE)}.\n`);
+    }
+}
+if (!isMainThread) {
+    if (workerData?.type === 'search_shard') {
+        const result = runSearchShard(workerData.payload, (p) => {
+            parentPort.postMessage({ type: 'progress', ...p });
+        });
+        parentPort.postMessage({ type: 'done', result });
+    }
+    else {
+        throw new Error('Unknown worker task type');
+    }
+}
+else if (require.main === module) {
+    run().catch((err) => {
+        console.error('Fatal:', err);
+        process.exit(1);
+    });
+}
+module.exports = {
+    parseArgs,
+};
+//# sourceMappingURL=optimizer_high_resolution.js.map