{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "1ea2b20e",
   "metadata": {},
   "source": [
    "# Régression logistique pour la génération de prénoms: corrigé\n",
    "\n",
    "Cours des grands algos d'hier et ajd, séance 1.\n",
    "  - pur python\n",
    "  - regression logistique par descente de gradient, calcul explicite du gradient."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "e5700c80",
   "metadata": {},
   "outputs": [],
   "source": [
    "import math\n",
    "import random\n",
    "random.seed(42)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d4a84756",
   "metadata": {},
   "source": [
    "## Le jeu de données\n",
    "\n",
    "Une liste de prénoms (un par ligne)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "cac02fe2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "num names: 32033\n",
      "['yuheng', 'diondre', 'xavien', 'jori', 'juanluis', 'erandi', 'phia', 'samatha']\n"
     ]
    }
   ],
   "source": [
    "import os\n",
    "if not os.path.exists('input.txt'):\n",
    "    import urllib.request\n",
    "    names_url = 'https://raw.githubusercontent.com/karpathy/makemore/988aa59/names.txt'\n",
    "    urllib.request.urlretrieve(names_url, 'input.txt')\n",
    "names = [line.strip() for line in open('input.txt') if line.strip()]\n",
    "random.shuffle(names)\n",
    "print(f'num names: {len(names)}')\n",
    "print(names[:8])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "198fe5f7",
   "metadata": {},
   "source": [
    "## Le tokeniseur\n",
    "\n",
    "Les tokens sont les caractères de l'alphabet, plus un token spécial `stop_token` (début/fin de mot).\n",
    "Il y a donc `T = 27` tokens."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "92a18b7b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "vocab size: T = 27\n",
      "anna -> [26, 0, 13, 13, 0, 26]\n"
     ]
    }
   ],
   "source": [
    "characters = sorted(set(''.join(names)))\n",
    "stop_token = len(characters)\n",
    "T = len(characters) + 1\n",
    "print(f'vocab size: T = {T}')\n",
    "\n",
    "def encode(name):\n",
    "    return [stop_token] + [characters.index(char) for char in name] + [stop_token]\n",
    "\n",
    "print('anna ->', encode('anna'))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e49298e4",
   "metadata": {},
   "source": [
    "## 3. Algèbre linéaire en Python pur\n",
    "\n",
    "Un vecteur = une liste de flottants ; une matrice = une liste de vecteurs lignes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e64fed4d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# vecteurs\n",
    "rand_vec = lambda d, std=1: [random.gauss(0, std) for _ in range(d)]\n",
    "one_hot = lambda t, T: [1.0 if i == t else 0.0 for i in range(T)]\n",
    "ext_vec = lambda alpha, u: [alpha * ui for ui in u]\n",
    "dot_prod = lambda u, v: sum(ui * vi for ui, vi in zip(u, v))\n",
    "vec_add = lambda u, v: [ui + vi for ui, vi in zip(u, v)]\n",
    "norm = lambda u: math.sqrt(dot_prod(u, u))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "c67ef087-6076-4690-b3b4-5b49d9d0c0c6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "rand_vec(3, 1)=[-0.5341476169657658, 0.9620173494831367, 1.3555262632633958]\n",
      "one_hot(1,3)=[0.0, 1.0, 0.0]\n",
      "ext_vec(0.5, v)=[-0.5, -0.5]\n",
      "dot_prod(u,v)=-2.0, dot_prod(u,u)=2.0\n",
      "vec_add(u,v)=[0.0, 0.0]\n",
      "norm(u)=1.4142135623730951\n"
     ]
    }
   ],
   "source": [
    "# tests\n",
    "u, v = [1., 1.], [-1.,-1.]\n",
    "print(f\"{rand_vec(3, 1)=}\")\n",
    "print(f\"{one_hot(1,3)=}\")\n",
    "print(f\"{ext_vec(0.5, v)=}\")\n",
    "print(f\"{dot_prod(u,v)=}, {dot_prod(u,u)=}\")\n",
    "print(f\"{vec_add(u,v)=}\")\n",
    "print(f\"{norm(u)=}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "c13429d3-ea46-48ce-acd5-dcef13b828a1",
   "metadata": {},
   "outputs": [],
   "source": [
    "# matrices\n",
    "rand_mat = lambda n, m, std=1: [rand_vec(m, std) for _ in range(n)]\n",
    "transpose = lambda M: list(zip(*M))\n",
    "mat_vec  = lambda M, v: [dot_prod(mi_, v) for mi_ in M]\n",
    "mat_mat  = lambda A, B: [[dot_prod(ai_, b_j) for b_j in transpose(B)] for ai_ in A]\n",
    "# mat_mat  = lambda A, B: [[dot_prod(ai_, b_j) for ai_ in A] for b_j in transpose(B)]\n",
    "\n",
    "scal_mat = lambda a, M: [[a * mij for mij in mi_] for mi_ in M]\n",
    "mat_add  = lambda A, B: [vec_add(ai_, bi_) for ai_, bi_ in zip(A, B)]\n",
    "col_mat = lambda v: [[vi] for vi in v]\n",
    "row_mat = lambda v: [v]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "545f7a3f-f857-4ef5-a6dc-66c203cd1cf5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "transpose(M1)=[(0.0, 0.0), (1.0, 0.0)]\n",
      "mat_vec(I, v)=[-1.0, -1.0]\n",
      "mat_mat(I,M1)=[[0.0, 1.0], [0.0, 0.0]]\n",
      "mat_mat(M1,M1)=[[0.0, 0.0], [0.0, 0.0]]\n",
      "mat_mat(col_mat(ones), row_mat(ones))=[[1.0, 1.0], [1.0, 1.0]]\n"
     ]
    }
   ],
   "source": [
    "# tests\n",
    "I = [[1., 0.], [0.,1.]]\n",
    "M1 = [[0., 1.], [0.,0.]]\n",
    "ones = [1., 1.]\n",
    "print(f\"{transpose(M1)=}\")\n",
    "print(f\"{mat_vec(I, v)=}\")\n",
    "print(f\"{mat_mat(I,M1)=}\")\n",
    "print(f\"{mat_mat(M1,M1)=}\")\n",
    "print(f\"{mat_mat(col_mat(ones), row_mat(ones))=}\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "51658fdf",
   "metadata": {},
   "source": [
    "## 4. La représentation du contexte (feature)\n",
    "\n",
    "On résume le contexte $(t_0,\\dots,t_{l-1})$ par la **moyenne** des one-hot des tokens vus :\n",
    "$$ v = \\frac{1}{l}\\sum_{k=0}^{l-1} \\mathbf{1}_{t_k} \\in \\mathbb{R}^T. $$\n",
    "La coordonnée $m$ de $v$ est la fréquence du token $m$ dans le contexte."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "8945cfc0",
   "metadata": {},
   "outputs": [],
   "source": [
    "def features(context):\n",
    "    v = one_hot(context[0], T)\n",
    "    for t in context[1:]:\n",
    "        v = vec_add(v, one_hot(t, T))\n",
    "    return ext_vec(1 / len(context), v)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b7b46089",
   "metadata": {},
   "source": [
    "## 5. Le modèle : softmax(M v)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "d52ba141",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.25, 0.25, 0.25, 0.25]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# fonction softmax\n",
    "def softmax(z):\n",
    "    m = max(z) # numerical stability\n",
    "    exp_z = [math.exp(zi - m) for zi in z]\n",
    "    return ext_vec(1 / sum(exp_z), exp_z)\n",
    "softmax([0., 0., 0., 0.])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "d296546c",
   "metadata": {},
   "outputs": [],
   "source": [
    "def generate_name(M, max_chars=100):\n",
    "    context = [stop_token]\n",
    "    sample = []\n",
    "    for _ in range(max_chars):\n",
    "        v = features(context)\n",
    "        z = mat_vec(M, v)\n",
    "        p = softmax([zi for zi in z])\n",
    "        token = random.choices(range(T), weights=p)[0]\n",
    "        if token == stop_token:\n",
    "            break\n",
    "        sample.append(characters[token])\n",
    "        context.append(token)\n",
    "    return ''.join(sample)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "be3d154c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['raxwrgvbzeuaaebzljrzyaiweucvalaigtpa',\n",
       " 'kyftdddb',\n",
       " 'g',\n",
       " 'drbkrcrqbcugigcgeek',\n",
       " 'vetgwlhdinivqlahnlhutpjlmsqxgroqemgkujekzkiwhukevxnqlihvjgwsexuzyktsjpssluxpxuefyukyecmemrzzuoghzhon',\n",
       " 'lsavkohmrquuvccxsncynfmaavmkmewacolnotpdtpd',\n",
       " '',\n",
       " 'kfuvlerwu',\n",
       " 'rksjkyfcdzxpyakpnr',\n",
       " 'crskhmqmslygiqazcjuaktphpucghhmrwrfsocbpvhtzxnfbgocahrprubiyeyepxxykoa']"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# pas d'entrainement, on ne s'attend pas a un miracle\n",
    "# random.seed(42)\n",
    "M = [[0]*T]*T # matrice nulle, probas uniformes\n",
    "n_names = 10\n",
    "gen_names = [generate_name(M) for _ in range(n_names)]\n",
    "gen_names"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "01bb31e6",
   "metadata": {},
   "source": [
    "## 6. Le gradient explicite (comme dans le cours)\n",
    "\n",
    "Pour un token, forward puis backward **à la main** :\n",
    "\n",
    "- forward : $z = Mv$, $p = \\mathrm{softmax}(z)$, loss $= -\\log p_{t_l}$ ;\n",
    "- backward : $\\nabla_z \\mathcal{L} = p - \\mathbf{1}_{t_l}$, puis\n",
    "  $\\nabla_M \\mathcal{L} = (p - \\mathbf{1}_{t_l})\\, v^{\\top}$ (produit extérieur).\n",
    "\n",
    "La fonction ci-dessous traite **un prénom** : elle accumule la loss et le gradient\n",
    "sur tous ses tokens, et renvoie leur moyenne."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "84badf34",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3.295836866004329, 3.295836866004329)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def loss_and_grad(tokens, M):\n",
    "    grad_prenom = [[0.0] * T] * T   # accumulateur du gradient, meme forme que M\n",
    "    total_loss = 0.0\n",
    "    n_tok = 0\n",
    "    for i in range(len(tokens) - 1):\n",
    "        t0___tlm1 = tokens[:i + 1]\n",
    "        tl  = tokens[i + 1]\n",
    "        v = features(t0___tlm1)\n",
    "        z = mat_vec(M, v)\n",
    "        p = softmax(z)\n",
    "        total_loss += -math.log(p[tl])\n",
    "        n_tok += 1\n",
    "        p_m_onehot = p\n",
    "        p_m_onehot[tl] -= 1.0\n",
    "        grad_tok = mat_mat(col_mat(p_m_onehot), row_mat(v))\n",
    "        grad_prenom = mat_add(grad_prenom, grad_tok)\n",
    "    # moyenne sur les tokens du prenom\n",
    "    return total_loss / n_tok, scal_mat(1 / n_tok, grad_prenom)\n",
    "\n",
    "M = [[0]*T]*T\n",
    "loss_and_grad(encode('tommy'), M)[0], math.log(T)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1281081c",
   "metadata": {},
   "source": [
    "## 7. Entraînement : descente de gradient stochastique\n",
    "\n",
    "On tire un prénoms à chaque étape (ici `batch_size = 1`,\n",
    "soit la SGD puis on met à jour\n",
    "$M \\leftarrow M - \\eta\\, \\nabla_M \\mathcal{L}$ avec un pas $\\eta$ fixe.\n",
    "On enregistre la loss pour tracer la courbe."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "9d4068d6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "step  1000 / 1000 | loss 2.6635 (avg 2.8069)\n",
      "done in 1.46s\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "\n",
    "M = [[0]*T]*T\n",
    "random.seed(42)\n",
    "start_time = time.time()\n",
    "learning_rate = 1.0\n",
    "num_steps = 1000\n",
    "loss_history = []\n",
    "\n",
    "for step in range(num_steps):\n",
    "    tokens = encode(names[step])\n",
    "    loss, grad = loss_and_grad(tokens, M)\n",
    "    avg_loss = loss if step == 0 else (avg_loss * step + loss) / (step + 1)\n",
    "    M = mat_add(M, scal_mat(-learning_rate, grad))      # M <- M - lr * grad\n",
    "    loss_history.append(loss)\n",
    "    if step % 250 == 0 or step == num_steps - 1:\n",
    "        print(f'step {step + 1:5d} / {num_steps} | loss {loss:.4f} (avg {avg_loss:.4f})', end='\\r')\n",
    "print(f\"\\ndone in {time.time() - start_time:.2f}s\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a65d1324",
   "metadata": {},
   "source": [
    "## 8. La courbe d'entraînement\n",
    "\n",
    "La loss brute (par prénom) est bruitée, ce n'est qu'un terme dans la fonction de coût qui est la moyenne sur tous les prénoms. On cacule une moyenne glissante pour lisser. Elle doit partir près de $\\log T \\approx 3.30$ et décroître."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "108b3660",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def moving_average(x, k=100):\n",
    "    out, s = [], 0.0\n",
    "    from collections import deque\n",
    "    win = deque()\n",
    "    for xi in x:\n",
    "        win.append(xi); s += xi\n",
    "        if len(win) > k: s -= win.popleft()\n",
    "        out.append(s / len(win))\n",
    "    return out\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "plt.figure(figsize=(8, 4))\n",
    "plt.plot(loss_history, alpha=0.3, label='loss (par prenom)')\n",
    "plt.plot(moving_average(loss_history, num_steps // 100), label='moyenne glissante (100)')\n",
    "plt.axhline(math.log(T), color='gray', ls='--', label='log(T) (uniforme)')\n",
    "plt.xlabel('etape (SGD)')\n",
    "plt.ylabel('log-vraisemblance negative')\n",
    "plt.title('Courbe d\\'entrainement — regression logistique (gradient explicite)')\n",
    "plt.legend()\n",
    "plt.tight_layout()\n",
    "plt.ylim(2, 3.5)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "260e8d23",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(['ali',\n",
       "  'caljeg',\n",
       "  'itna',\n",
       "  'rzywclejaj',\n",
       "  'crii',\n",
       "  'zkobf',\n",
       "  'kjpanmyc',\n",
       "  'aeosie',\n",
       "  'bnokon',\n",
       "  'arfjqhivan',\n",
       "  'zaei',\n",
       "  'sthaytq',\n",
       "  'mathr',\n",
       "  'acaniy',\n",
       "  'anio',\n",
       "  'ralgayshyonab',\n",
       "  'sktmaai',\n",
       "  'rtydea',\n",
       "  'shna',\n",
       "  'szswash'],\n",
       " 6.25,\n",
       " 6.122217712983486)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gen_names = [generate_name(M) for _ in range(20)]\n",
    "gen_names,  sum([len(nm) for nm in gen_names])/ len(gen_names), sum([len(nm) for nm in names]) / len(names)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.12.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}
