@machinespirits/eval 0.1.2 → 0.2.0

package/services/anovaStats.js

@@ -0,0 +1,499 @@
+/**
+ * ANOVA Statistics Module
+ *
+ * Three-way ANOVA for 2×2×2 factorial designs.
+ * Extracted from benchmarkService to be reusable by the main evaluation pipeline.
+ *
+ * Factors:
+ *   A: Recognition (standard vs recognition-enhanced prompts)
+ *   B: Multi-agent tutor (single vs ego+superego dialogue)
+ *   C: Multi-agent learner (unified vs ego_superego)
+ */
+
+// ---- F-distribution p-value via regularized incomplete beta function ----
+
+/**
+ * Log-gamma function using Lanczos approximation (g=7, n=9 coefficients).
+ * Accurate to ~15 decimal digits for positive real arguments.
+ * Uses the reflection formula for z < 0.5.
+ */
+function lnGamma(z) {
+  if (z <= 0) return Infinity;
+
+  // Reflection formula: Gamma(z)*Gamma(1-z) = pi/sin(pi*z)
+  if (z < 0.5) {
+    return Math.log(Math.PI / Math.sin(Math.PI * z)) - lnGamma(1 - z);
+  }
+
+  const g = 7;
+  const c = [
+    0.99999999999980993,
+    676.5203681218851,
+    -1259.1392167224028,
+    771.32342877765313,
+    -176.61502916214059,
+    12.507343278686905,
+    -0.13857109526572012,
+    9.9843695780195716e-6,
+    1.5056327351493116e-7,
+  ];
+
+  z -= 1;
+  let x = c[0];
+  for (let i = 1; i < g + 2; i++) {
+    x += c[i] / (z + i);
+  }
+  const t = z + g + 0.5;
+  return 0.5 * Math.log(2 * Math.PI) + (z + 0.5) * Math.log(t) - t + Math.log(x);
+}
+
+/**
+ * Regularized incomplete beta function I_x(a, b) via the continued fraction
+ * representation from Numerical Recipes (betacf). Uses modified Lentz's method.
+ *
+ * The continued fraction is:
+ *   I_x(a,b) = prefactor * (1/1+) (d1/1+) (d2/1+) (d3/1+) ...
+ * where d_{2m+1} = -(a+m)(a+b+m)x / ((a+2m)(a+2m+1))
+ *       d_{2m}   = m(b-m)x / ((a+2m-1)(a+2m))
+ */
+function regularizedBeta(x, a, b) {
+  if (x <= 0) return 0;
+  if (x >= 1) return 1;
+
+  // Use the symmetry relation when x > (a+1)/(a+b+2) for better convergence
+  if (x > (a + 1) / (a + b + 2)) {
+    return 1 - regularizedBeta(1 - x, b, a);
+  }
+
+  // Compute the prefactor: x^a * (1-x)^b / (a * Beta(a,b))
+  const lnPrefactor = a * Math.log(x) + b * Math.log(1 - x)
+    - Math.log(a) - lnGamma(a) - lnGamma(b) + lnGamma(a + b);
+  const prefactor = Math.exp(lnPrefactor);
+
+  // Evaluate the continued fraction using modified Lentz's method
+  // Following Numerical Recipes "betacf" algorithm
+  const maxIter = 200;
+  const eps = 3e-14;
+  const fpmin = 1e-30;
+
+  let qab = a + b;
+  let qap = a + 1;
+  let qam = a - 1;
+  let c = 1;
+  let d = 1 - qab * x / qap;
+  if (Math.abs(d) < fpmin) d = fpmin;
+  d = 1 / d;
+  let h = d;
+
+  for (let m = 1; m <= maxIter; m++) {
+    // Even step: d_{2m}
+    let aa = m * (b - m) * x / ((qam + 2 * m) * (a + 2 * m));
+    d = 1 + aa * d;
+    if (Math.abs(d) < fpmin) d = fpmin;
+    c = 1 + aa / c;
+    if (Math.abs(c) < fpmin) c = fpmin;
+    d = 1 / d;
+    h *= d * c;
+
+    // Odd step: d_{2m+1}
+    aa = -(a + m) * (qab + m) * x / ((a + 2 * m) * (qap + 2 * m));
+    d = 1 + aa * d;
+    if (Math.abs(d) < fpmin) d = fpmin;
+    c = 1 + aa / c;
+    if (Math.abs(c) < fpmin) c = fpmin;
+    d = 1 / d;
+    const del = d * c;
+    h *= del;
+
+    if (Math.abs(del - 1) < eps) break;
+  }
+
+  return prefactor * h;
+}
+
+/**
+ * Compute p-value for the F-distribution: P(F > x | d1, d2).
+ *
+ * @param {number} F - The F-statistic
+ * @param {number} d1 - Numerator degrees of freedom
+ * @param {number} d2 - Denominator degrees of freedom
+ * @returns {number} p-value (upper tail probability)
+ */
+function fDistPValue(F, d1, d2) {
+  if (F <= 0 || d1 <= 0 || d2 <= 0) return 1;
+  if (!isFinite(F)) return 0;
+
+  const x = d1 * F / (d1 * F + d2);
+  return 1 - regularizedBeta(x, d1 / 2, d2 / 2);
+}
+
+/**
+ * Run a three-way ANOVA on factorial cell data.
+ *
+ * @param {Object} data - Map of cellKey → [scores]
+ *   Cell keys encode factor levels: "r{0|1}_t{0|1}_l{0|1}"
+ *   e.g. { "r0_t0_l0": [55, 62, 58], "r0_t0_l1": [60, 65, 63], ... }
+ * @returns {Object} ANOVA results with main effects, interactions, and diagnostics
+ */
+export function runThreeWayANOVA(data) {
+  const cells = {};
+
+  // Accept both cellKey format ("r0_t0_l0") and profile-name format
+  for (const [key, scores] of Object.entries(data)) {
+    if (key.match(/^r[01]_t[01]_l[01]$/)) {
+      cells[key] = scores;
+    }
+  }
+
+  // If no cell-keyed data, return error
+  if (Object.keys(cells).length === 0) {
+    return { error: 'No data available for ANOVA. Data must use cell keys like "r0_t0_l0".' };
+  }
+
+  const allData = Object.values(cells).flat();
+  const N = allData.length;
+  if (N === 0) {
+    return { error: 'No data available for ANOVA' };
+  }
+
+  const grandMean = allData.reduce((a, b) => a + b, 0) / N;
+
+  const getByFactors = (r, t, l) => cells[`r${r}_t${t}_l${l}`] || [];
+
+  // Calculate marginal means
+  const getMarginalMean = (factor, level) => {
+    let values = [];
+    if (factor === 'recognition') {
+      for (const t of [0, 1]) {
+        for (const l of [0, 1]) {
+          values = values.concat(getByFactors(level, t, l));
+        }
+      }
+    } else if (factor === 'tutor') {
+      for (const r of [0, 1]) {
+        for (const l of [0, 1]) {
+          values = values.concat(getByFactors(r, level, l));
+        }
+      }
+    } else if (factor === 'learner') {
+      for (const r of [0, 1]) {
+        for (const t of [0, 1]) {
+          values = values.concat(getByFactors(r, t, level));
+        }
+      }
+    }
+    return values.length > 0 ? values.reduce((a, b) => a + b, 0) / values.length : grandMean;
+  };
+
+  const meanR0 = getMarginalMean('recognition', 0);
+  const meanR1 = getMarginalMean('recognition', 1);
+  const meanT0 = getMarginalMean('tutor', 0);
+  const meanT1 = getMarginalMean('tutor', 1);
+  const meanL0 = getMarginalMean('learner', 0);
+  const meanL1 = getMarginalMean('learner', 1);
+
+  // Sample sizes per level
+  const getN = (factor, level) => {
+    let count = 0;
+    if (factor === 'recognition') {
+      for (const t of [0, 1]) {
+        for (const l of [0, 1]) {
+          count += getByFactors(level, t, l).length;
+        }
+      }
+    } else if (factor === 'tutor') {
+      for (const r of [0, 1]) {
+        for (const l of [0, 1]) {
+          count += getByFactors(r, level, l).length;
+        }
+      }
+    } else if (factor === 'learner') {
+      for (const r of [0, 1]) {
+        for (const t of [0, 1]) {
+          count += getByFactors(r, t, level).length;
+        }
+      }
+    }
+    return count;
+  };
+
+  // Sum of Squares
+  const SST = allData.reduce((acc, x) => acc + (x - grandMean) ** 2, 0);
+
+  const nR0 = getN('recognition', 0);
+  const nR1 = getN('recognition', 1);
+  const nT0 = getN('tutor', 0);
+  const nT1 = getN('tutor', 1);
+  const nL0 = getN('learner', 0);
+  const nL1 = getN('learner', 1);
+
+  const SS_R = nR0 * (meanR0 - grandMean) ** 2 + nR1 * (meanR1 - grandMean) ** 2;
+  const SS_T = nT0 * (meanT0 - grandMean) ** 2 + nT1 * (meanT1 - grandMean) ** 2;
+  const SS_L = nL0 * (meanL0 - grandMean) ** 2 + nL1 * (meanL1 - grandMean) ** 2;
+
+  // Two-way interaction means
+  const getTwoWayMean = (f1, l1, f2, l2) => {
+    let values = [];
+    if (f1 === 'recognition' && f2 === 'tutor') {
+      for (const l of [0, 1]) values = values.concat(getByFactors(l1, l2, l));
+    } else if (f1 === 'recognition' && f2 === 'learner') {
+      for (const t of [0, 1]) values = values.concat(getByFactors(l1, t, l2));
+    } else if (f1 === 'tutor' && f2 === 'learner') {
+      for (const r of [0, 1]) values = values.concat(getByFactors(r, l1, l2));
+    }
+    return values.length > 0 ? values.reduce((a, b) => a + b, 0) / values.length : grandMean;
+  };
+
+  let SS_RT = 0, SS_RL = 0, SS_TL = 0;
+  for (const r of [0, 1]) {
+    for (const t of [0, 1]) {
+      const cellMean = getTwoWayMean('recognition', r, 'tutor', t);
+      const expected = (r === 1 ? meanR1 : meanR0) + (t === 1 ? meanT1 : meanT0) - grandMean;
+      const cellN = getByFactors(r, t, 0).length + getByFactors(r, t, 1).length;
+      SS_RT += cellN * (cellMean - expected) ** 2;
+    }
+  }
+  for (const r of [0, 1]) {
+    for (const l of [0, 1]) {
+      const cellMean = getTwoWayMean('recognition', r, 'learner', l);
+      const expected = (r === 1 ? meanR1 : meanR0) + (l === 1 ? meanL1 : meanL0) - grandMean;
+      const cellN = getByFactors(r, 0, l).length + getByFactors(r, 1, l).length;
+      SS_RL += cellN * (cellMean - expected) ** 2;
+    }
+  }
+  for (const t of [0, 1]) {
+    for (const l of [0, 1]) {
+      const cellMean = getTwoWayMean('tutor', t, 'learner', l);
+      const expected = (t === 1 ? meanT1 : meanT0) + (l === 1 ? meanL1 : meanL0) - grandMean;
+      const cellN = getByFactors(0, t, l).length + getByFactors(1, t, l).length;
+      SS_TL += cellN * (cellMean - expected) ** 2;
+    }
+  }
+
+  // Three-way interaction
+  let SS_cells = 0;
+  for (const r of [0, 1]) {
+    for (const t of [0, 1]) {
+      for (const l of [0, 1]) {
+        const cellData = getByFactors(r, t, l);
+        if (cellData.length > 0) {
+          const cellMean = cellData.reduce((a, b) => a + b, 0) / cellData.length;
+          SS_cells += cellData.length * (cellMean - grandMean) ** 2;
+        }
+      }
+    }
+  }
+  const SS_RTL = Math.max(0, SS_cells - SS_R - SS_T - SS_L - SS_RT - SS_RL - SS_TL);
+
+  // Error SS (within cells)
+  let SS_E = 0;
+  for (const r of [0, 1]) {
+    for (const t of [0, 1]) {
+      for (const l of [0, 1]) {
+        const cellData = getByFactors(r, t, l);
+        if (cellData.length > 0) {
+          const cellMean = cellData.reduce((a, b) => a + b, 0) / cellData.length;
+          SS_E += cellData.reduce((acc, x) => acc + (x - cellMean) ** 2, 0);
+        }
+      }
+    }
+  }
+
+  // Degrees of freedom
+  const df_R = 1, df_T = 1, df_L = 1;
+  const df_RT = 1, df_RL = 1, df_TL = 1;
+  const df_RTL = 1;
+  const df_E = N - 8;
+  const df_T_total = N - 1;
+
+  // Mean Squares
+  const MS_R = SS_R / df_R;
+  const MS_T = SS_T / df_T;
+  const MS_L = SS_L / df_L;
+  const MS_RT = SS_RT / df_RT;
+  const MS_RL = SS_RL / df_RL;
+  const MS_TL = SS_TL / df_TL;
+  const MS_RTL = SS_RTL / df_RTL;
+  const MS_E = df_E > 0 ? SS_E / df_E : 1;
+
+  // F ratios
+  const F_R = MS_R / MS_E;
+  const F_T = MS_T / MS_E;
+  const F_L = MS_L / MS_E;
+  const F_RT = MS_RT / MS_E;
+  const F_RL = MS_RL / MS_E;
+  const F_TL = MS_TL / MS_E;
+  const F_RTL = MS_RTL / MS_E;
+
+  // Compute p-values from the F distribution CDF
+  const getP = (F, df1, df2) => fDistPValue(F, df1, df2);
+
+  const etaSq = (SS) => SST > 0 ? SS / SST : 0;
+
+  return {
+    grandMean,
+    N,
+    marginalMeans: {
+      recognition: { standard: meanR0, recognition: meanR1 },
+      tutor: { single: meanT0, multi: meanT1 },
+      learner: { unified: meanL0, ego_superego: meanL1 },
+    },
+    mainEffects: {
+      recognition: { SS: SS_R, df: df_R, MS: MS_R, F: F_R, p: getP(F_R, df_R, df_E), etaSq: etaSq(SS_R) },
+      tutor: { SS: SS_T, df: df_T, MS: MS_T, F: F_T, p: getP(F_T, df_T, df_E), etaSq: etaSq(SS_T) },
+      learner: { SS: SS_L, df: df_L, MS: MS_L, F: F_L, p: getP(F_L, df_L, df_E), etaSq: etaSq(SS_L) },
+    },
+    interactions: {
+      recognition_x_tutor: { SS: SS_RT, df: df_RT, MS: MS_RT, F: F_RT, p: getP(F_RT, df_RT, df_E), etaSq: etaSq(SS_RT) },
+      recognition_x_learner: { SS: SS_RL, df: df_RL, MS: MS_RL, F: F_RL, p: getP(F_RL, df_RL, df_E), etaSq: etaSq(SS_RL) },
+      tutor_x_learner: { SS: SS_TL, df: df_TL, MS: MS_TL, F: F_TL, p: getP(F_TL, df_TL, df_E), etaSq: etaSq(SS_TL) },
+      three_way: { SS: SS_RTL, df: df_RTL, MS: MS_RTL, F: F_RTL, p: getP(F_RTL, df_RTL, df_E), etaSq: etaSq(SS_RTL) },
+    },
+    error: { SS: SS_E, df: df_E, MS: MS_E },
+    total: { SS: SST, df: df_T_total },
+  };
+}
+
+/**
+ * Convert profile factor tags to ANOVA cell keys.
+ *
+ * @param {Object} factors - { recognition: bool, multi_agent_tutor: bool, multi_agent_learner: bool }
+ * @returns {string} Cell key like "r0_t1_l0"
+ */
+export function factorsToCellKey(factors) {
+  const r = factors.recognition ? 1 : 0;
+  const t = factors.multi_agent_tutor ? 1 : 0;
+  const l = factors.multi_agent_learner ? 1 : 0;
+  return `r${r}_t${t}_l${l}`;
+}
+
+/**
+ * Format ANOVA results as a text report.
+ *
+ * @param {Object} anovaResults - Output of runThreeWayANOVA()
+ * @param {Object} [options] - Formatting options
+ * @param {string} [options.scoreLabel] - Label for the score type (e.g. "Base Score", "Recognition Score")
+ * @returns {string} Formatted report
+ */
+export function formatANOVAReport(anovaResults, options = {}) {
+  const { scoreLabel } = options;
+
+  if (typeof anovaResults.error === 'string') {
+    return `ANOVA Error: ${anovaResults.error}`;
+  }
+
+  const lines = [];
+  const title = scoreLabel
+    ? `THREE-WAY ANOVA: ${scoreLabel.toUpperCase()}`
+    : 'THREE-WAY ANOVA: 2x2x2 FACTORIAL ANALYSIS';
+
+  lines.push('');
+  lines.push('='.repeat(70));
+  lines.push(`  ${title}`);
+  lines.push('='.repeat(70));
+  lines.push(`  Grand Mean: ${anovaResults.grandMean.toFixed(2)}  |  N = ${anovaResults.N}`);
+  lines.push('');
+
+  // Marginal means
+  const mm = anovaResults.marginalMeans;
+  lines.push('-'.repeat(70));
+  lines.push('  MARGINAL MEANS');
+  lines.push('-'.repeat(70));
+  lines.push(`  Recognition:   Standard = ${mm.recognition.standard.toFixed(2)},  Recognition = ${mm.recognition.recognition.toFixed(2)}`);
+  lines.push(`  Tutor:         Single = ${mm.tutor.single.toFixed(2)},  Multi-Agent = ${mm.tutor.multi.toFixed(2)}`);
+  lines.push(`  Learner:       Unified = ${mm.learner.unified.toFixed(2)},  Ego/Superego = ${mm.learner.ego_superego.toFixed(2)}`);
+  lines.push('');
+
+  // ANOVA table
+  lines.push('-'.repeat(70));
+  lines.push('  ANOVA TABLE');
+  lines.push('-'.repeat(70));
+  lines.push('  Source                    SS       df       MS        F        p       eta2');
+  lines.push('  ' + '-'.repeat(66));
+
+  const formatRow = (name, data) => {
+    const ss = data.SS.toFixed(2).padStart(8);
+    const df = data.df.toString().padStart(6);
+    const ms = data.MS.toFixed(2).padStart(8);
+    const f = data.F.toFixed(3).padStart(8);
+    const p = data.p < 0.001 ? '< .001' : data.p.toFixed(3);
+    const eta = data.etaSq.toFixed(3).padStart(6);
+    const sig = data.p < 0.05 ? '***' : (data.p < 0.1 ? '*' : '');
+    return `  ${name.padEnd(22)}  ${ss}  ${df}  ${ms}  ${f}  ${p.padStart(8)}  ${eta}  ${sig}`;
+  };
+
+  const me = anovaResults.mainEffects;
+  const ia = anovaResults.interactions;
+
+  lines.push(formatRow('Recognition (A)', me.recognition));
+  lines.push(formatRow('Tutor Architecture (B)', me.tutor));
+  lines.push(formatRow('Learner Arch. (C)', me.learner));
+  lines.push('  ' + '-'.repeat(66));
+  lines.push(formatRow('A x B', ia.recognition_x_tutor));
+  lines.push(formatRow('A x C', ia.recognition_x_learner));
+  lines.push(formatRow('B x C', ia.tutor_x_learner));
+  lines.push(formatRow('A x B x C', ia.three_way));
+  lines.push('  ' + '-'.repeat(66));
+
+  const err = anovaResults.error;
+  lines.push(`  ${'Error'.padEnd(22)}  ${err.SS.toFixed(2).padStart(8)}  ${err.df.toString().padStart(6)}  ${err.MS.toFixed(2).padStart(8)}`);
+  lines.push('');
+  lines.push('  Significance: *** p < .05, * p < .10');
+  lines.push('');
+
+  // Interpretation
+  lines.push('-'.repeat(70));
+  lines.push('  INTERPRETATION');
+  lines.push('-'.repeat(70));
+
+  const formatP = (p) => p < 0.001 ? '< .001' : `= .${p.toFixed(3).slice(2)}`;
+
+  if (me.recognition.p < 0.05) {
+    const effect = mm.recognition.recognition - mm.recognition.standard;
+    lines.push(`  * Recognition prompts: SIGNIFICANT (F = ${me.recognition.F.toFixed(2)}, p ${formatP(me.recognition.p)})`);
+    lines.push(`    Effect: ${effect >= 0 ? '+' : ''}${effect.toFixed(2)} points, eta2 = ${me.recognition.etaSq.toFixed(3)}`);
+  } else {
+    lines.push(`  - Recognition prompts: not significant (F = ${me.recognition.F.toFixed(2)}, p ${formatP(me.recognition.p)})`);
+  }
+
+  if (me.tutor.p < 0.05) {
+    const effect = mm.tutor.multi - mm.tutor.single;
+    lines.push(`  * Multi-agent tutor: SIGNIFICANT (F = ${me.tutor.F.toFixed(2)}, p ${formatP(me.tutor.p)})`);
+    lines.push(`    Effect: ${effect >= 0 ? '+' : ''}${effect.toFixed(2)} points, eta2 = ${me.tutor.etaSq.toFixed(3)}`);
+  } else {
+    lines.push(`  - Multi-agent tutor: not significant (F = ${me.tutor.F.toFixed(2)}, p ${formatP(me.tutor.p)})`);
+  }
+
+  if (me.learner.p < 0.05) {
+    const effect = mm.learner.ego_superego - mm.learner.unified;
+    lines.push(`  * Multi-agent learner: SIGNIFICANT (F = ${me.learner.F.toFixed(2)}, p ${formatP(me.learner.p)})`);
+    lines.push(`    Effect: ${effect >= 0 ? '+' : ''}${effect.toFixed(2)} points, eta2 = ${me.learner.etaSq.toFixed(3)}`);
+  } else {
+    lines.push(`  - Multi-agent learner: not significant (F = ${me.learner.F.toFixed(2)}, p ${formatP(me.learner.p)})`);
+  }
+
+  // Interactions
+  lines.push('');
+  const interactions = [
+    { key: 'recognition_x_tutor', label: 'Recognition x Tutor' },
+    { key: 'recognition_x_learner', label: 'Recognition x Learner' },
+    { key: 'tutor_x_learner', label: 'Tutor x Learner' },
+    { key: 'three_way', label: 'Three-way' },
+  ];
+  for (const { key, label } of interactions) {
+    if (ia[key].p < 0.05) {
+      lines.push(`  * ${label} interaction: SIGNIFICANT (F = ${ia[key].F.toFixed(2)}, p ${formatP(ia[key].p)})`);
+    }
+  }
+
+  lines.push('');
+  lines.push('='.repeat(70));
+
+  return lines.join('\n');
+}
+
+export default {
+  runThreeWayANOVA,
+  factorsToCellKey,
+  formatANOVAReport,
+};