{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "14235502",
   "metadata": {},
   "source": [
    "# Marketing mix modeling from scratch\n",
    "\n",
    "This notebook rebuilds the model behind the free [Marketing Mix Modeling Lab](https://realanalystjobs.com/learn/marketing-mix-modeling) with pandas and NumPy only, step by step:\n",
    "\n",
    "1. Load weekly data\n",
    "2. Adstock (carryover) and Hill saturation (diminishing returns)\n",
    "3. Ridge regression with non-negative channel effects\n",
    "4. Search the adstock and saturation settings on a validation window\n",
    "5. Check the model on a test window it never saw\n",
    "6. Bootstrap for ranges\n",
    "7. ROI, next-dollar return, and a check against the answer key\n",
    "\n",
    "**No Jupyter installed?** Open [Google Colab](https://colab.research.google.com), choose *File > Upload notebook*, pick this file, then upload `mmm-synthetic-example.csv` with the folder icon on the left. Everything here also runs in Colab as is.\n",
    "\n",
    "The data is **synthetic**: it was simulated from known settings, so at the end you can grade the model.\n",
    "\n",
    "**Expect different numbers from the lab.** The lab pools four independent searches, 160 bootstrap refits and a calibrated widening of its ranges. This notebook runs one smaller search and 40 refits so it finishes in under a minute, so a channel's ROI or half-saturation can differ noticeably from the lab's. That is not a bug in your code: judge the notebook by the answer key at the end, not by the lab."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "2f3333ca",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-11-10T05:50:24.401698Z",
     "iopub.status.busy": "2026-11-10T05:50:24.401448Z",
     "iopub.status.idle": "2026-11-10T05:50:25.109410Z",
     "shell.execute_reply": "2026-11-10T05:50:25.108736Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>date</th>\n",
       "      <th>revenue</th>\n",
       "      <th>google_search_spend</th>\n",
       "      <th>meta_spend</th>\n",
       "      <th>offline_tv_spend</th>\n",
       "      <th>promo</th>\n",
       "      <th>holiday</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2023-01-02</td>\n",
       "      <td>244055</td>\n",
       "      <td>22967</td>\n",
       "      <td>13864</td>\n",
       "      <td>19175</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2023-01-09</td>\n",
       "      <td>261940</td>\n",
       "      <td>22876</td>\n",
       "      <td>14620</td>\n",
       "      <td>20020</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2023-01-16</td>\n",
       "      <td>256627</td>\n",
       "      <td>24516</td>\n",
       "      <td>4080</td>\n",
       "      <td>15290</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2023-01-23</td>\n",
       "      <td>287278</td>\n",
       "      <td>37854</td>\n",
       "      <td>14003</td>\n",
       "      <td>34260</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2023-01-30</td>\n",
       "      <td>285705</td>\n",
       "      <td>20929</td>\n",
       "      <td>16326</td>\n",
       "      <td>27981</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        date  revenue  google_search_spend  meta_spend  offline_tv_spend  \\\n",
       "0 2023-01-02   244055                22967       13864             19175   \n",
       "1 2023-01-09   261940                22876       14620             20020   \n",
       "2 2023-01-16   256627                24516        4080             15290   \n",
       "3 2023-01-23   287278                37854       14003             34260   \n",
       "4 2023-01-30   285705                20929       16326             27981   \n",
       "\n",
       "   promo  holiday  \n",
       "0      0        0  \n",
       "1      0        0  \n",
       "2      0        0  \n",
       "3      0        0  \n",
       "4      0        0  "
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import warnings\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# NumPy 2.0 on some Macs prints spurious matmul overflow warnings; the results are unaffected.\n",
    "warnings.filterwarnings(\"ignore\", message=\".*encountered in matmul\")\n",
    "rng = np.random.default_rng(42)\n",
    "# Download the CSV from the lab page, or point this at your own weekly data.\n",
    "df = pd.read_csv(\"mmm-synthetic-example.csv\", comment=\"#\", parse_dates=[\"date\"])\n",
    "channels = [\"google_search_spend\", \"meta_spend\", \"offline_tv_spend\"]\n",
    "controls = [\"promo\", \"holiday\"]\n",
    "kpi = \"revenue\"\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b291e015",
   "metadata": {},
   "source": [
    "## 1. The two transforms\n",
    "\n",
    "**Adstock**: this week's pressure is this week's spend plus `decay` times last week's pressure, so an ad keeps working after the week it runs.\n",
    "\n",
    "**Hill saturation**: `x^s / (x^s + h^s)`, where `x` is adstocked spend divided by its highest week. Returns bend down once `x` passes about `h`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "75a858b8",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-11-10T05:50:25.112340Z",
     "iopub.status.busy": "2026-11-10T05:50:25.112043Z",
     "iopub.status.idle": "2026-11-10T05:50:25.308567Z",
     "shell.execute_reply": "2026-11-10T05:50:25.307987Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x320 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def adstock(x, decay):\n",
    "    out = np.zeros(len(x))\n",
    "    carry = 0.0\n",
    "    for i, v in enumerate(x):\n",
    "        carry = v + decay * carry\n",
    "        out[i] = carry\n",
    "    return out\n",
    "\n",
    "def hill(x, half_sat, shape):\n",
    "    x = np.maximum(x, 0)\n",
    "    return x**shape / (x**shape + half_sat**shape)\n",
    "\n",
    "# One unit of spend in week 0: how it lands over the next weeks, for three decays\n",
    "fig, ax = plt.subplots(1, 2, figsize=(10, 3.2))\n",
    "for d in (0.2, 0.5, 0.8):\n",
    "    ax[0].plot(adstock(np.r_[1, np.zeros(9)], d), marker=\"o\", label=f\"decay {d}\")\n",
    "ax[0].set(title=\"Adstock: one week of spend carries forward\", xlabel=\"weeks after\", ylabel=\"pressure\")\n",
    "ax[0].legend()\n",
    "x = np.linspace(0, 1.5, 100)\n",
    "for h_, s_ in ((0.3, 1.0), (0.5, 2.0), (0.7, 3.0)):\n",
    "    ax[1].plot(x, hill(x, h_, s_), label=f\"h={h_}, s={s_}\")\n",
    "ax[1].set(title=\"Hill: diminishing returns\", xlabel=\"adstocked spend / highest week\", ylabel=\"share of max effect\")\n",
    "ax[1].legend()\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "431c0fc7",
   "metadata": {},
   "source": [
    "## 2. The design matrix and ridge regression\n",
    "\n",
    "Baseline columns: a linear trend, two yearly sine and cosine waves, and the controls. Media columns: `hill(adstock(spend) / scale)`.\n",
    "\n",
    "Ridge regression shrinks coefficients a little to stay stable when columns are related. Channel effects are held at zero or above, because spending on ads should not lower sales. Coordinate descent makes that constraint easy."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "4995e46b",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-11-10T05:50:25.311254Z",
     "iopub.status.busy": "2026-11-10T05:50:25.311045Z",
     "iopub.status.idle": "2026-11-10T05:50:25.319511Z",
     "shell.execute_reply": "2026-11-10T05:50:25.318861Z"
    }
   },
   "outputs": [],
   "source": [
    "PERIOD = 52.1775\n",
    "\n",
    "def baseline_columns(t, n, ctrl):\n",
    "    cols = [t / max(1, n - 1)]\n",
    "    for k in (1, 2):\n",
    "        w = 2 * np.pi * k * t / PERIOD\n",
    "        cols += [np.sin(w), np.cos(w)]\n",
    "    return np.column_stack(cols + [ctrl])\n",
    "\n",
    "def design(data, hyper, scales=None):\n",
    "    n = len(data)\n",
    "    t = np.arange(n)\n",
    "    ctrl = data[controls].to_numpy(float)\n",
    "    ctrl = (ctrl - ctrl.mean(0)) / np.where(ctrl.std(0) > 0, ctrl.std(0), 1)\n",
    "    media, sc = [], []\n",
    "    for j, ch in enumerate(channels):\n",
    "        a = adstock(data[ch].to_numpy(float), hyper[j][0])\n",
    "        s = scales[j] if scales is not None else max(a.max(), 1e-9)\n",
    "        sc.append(s)\n",
    "        media.append(hill(a / s, hyper[j][1], hyper[j][2]))\n",
    "    return np.column_stack(media + [baseline_columns(t, n, ctrl)]), sc\n",
    "\n",
    "def ridge_nonneg(X, y, n_media, lam, iters=300):\n",
    "    mu, sd = X.mean(0), X.std(0)\n",
    "    sd[sd == 0] = 1\n",
    "    Z = (X - mu) / sd\n",
    "    n, p = Z.shape\n",
    "    b = np.zeros(p)\n",
    "    res = y - y.mean()\n",
    "    for _ in range(iters):\n",
    "        delta = 0.0\n",
    "        for j in range(p):\n",
    "            rho = Z[:, j] @ res + n * b[j]\n",
    "            nb = rho / (n * (1 + lam))\n",
    "            if j < n_media and nb < 0:\n",
    "                nb = 0.0\n",
    "            d = nb - b[j]\n",
    "            if d:\n",
    "                res -= d * Z[:, j]\n",
    "                b[j] = nb\n",
    "                delta = max(delta, abs(d))\n",
    "        if delta < 1e-9 * (abs(y.mean()) + 1):\n",
    "            break\n",
    "    beta = b / sd\n",
    "    return y.mean() - beta @ mu, beta"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c826e97d",
   "metadata": {},
   "source": [
    "## 3. Search the settings, with a validation window and a separate test window\n",
    "\n",
    "Carryover and saturation cannot be solved for directly, so we try many random settings. Each is fitted on the early weeks and scored on a **validation window**. The last weeks form a **test window** that plays no part in choosing anything, so its error is an honest check."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "4b32c56e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-11-10T05:50:25.321794Z",
     "iopub.status.busy": "2026-11-10T05:50:25.321609Z",
     "iopub.status.idle": "2026-11-10T05:50:27.408865Z",
     "shell.execute_reply": "2026-11-10T05:50:27.408235Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>decay</th>\n",
       "      <th>half_sat</th>\n",
       "      <th>shape</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>google_search_spend</th>\n",
       "      <td>0.45</td>\n",
       "      <td>0.79</td>\n",
       "      <td>1.63</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>meta_spend</th>\n",
       "      <td>0.50</td>\n",
       "      <td>0.93</td>\n",
       "      <td>1.95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>offline_tv_spend</th>\n",
       "      <td>0.77</td>\n",
       "      <td>0.81</td>\n",
       "      <td>0.93</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     decay  half_sat  shape\n",
       "google_search_spend   0.45      0.79   1.63\n",
       "meta_spend            0.50      0.93   1.95\n",
       "offline_tv_spend      0.77      0.81   0.93"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n = len(df)\n",
    "H = max(8, round(n * 0.08))           # weeks in each window\n",
    "y_all = df[kpi].to_numpy(float)\n",
    "LAMBDAS = [1e-4, 1e-3, 1e-2, 3e-2, 0.1]\n",
    "BOUNDS = dict(decay=(0, 0.85), half=(0.25, 0.95), shape=(0.8, 3))\n",
    "\n",
    "def random_hyper():\n",
    "    return [(rng.uniform(*BOUNDS[\"decay\"]), rng.uniform(*BOUNDS[\"half\"]), rng.uniform(*BOUNDS[\"shape\"])) for _ in channels]\n",
    "\n",
    "def nrmse(a, p):\n",
    "    return np.sqrt(np.mean((a - p) ** 2)) / (a.max() - a.min())\n",
    "\n",
    "def score(hyper, data, y):\n",
    "    # fit on all but the last H weeks of `data`, score on those H weeks\n",
    "    X, sc = design(data, hyper)\n",
    "    m = len(data) - H\n",
    "    best = None\n",
    "    for lam in LAMBDAS:\n",
    "        b0, beta = ridge_nonneg(X[:m], y[:m], len(channels), lam)\n",
    "        pred = b0 + X @ beta\n",
    "        s = nrmse(y[:m], pred[:m]) + nrmse(y[m:], pred[m:])\n",
    "        if best is None or s < best[0]:\n",
    "            best = (s, lam)\n",
    "    return best\n",
    "\n",
    "select = df.iloc[: n - H]              # the test window is cut off here\n",
    "candidates = []\n",
    "for i in range(300):\n",
    "    h = random_hyper()\n",
    "    s, lam = score(h, select, y_all[: n - H])\n",
    "    candidates.append((s, lam, h))\n",
    "candidates.sort(key=lambda c: c[0])\n",
    "best_score, best_lam, best_hyper = candidates[0]\n",
    "pd.DataFrame(best_hyper, index=channels, columns=[\"decay\", \"half_sat\", \"shape\"]).round(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "861b3381",
   "metadata": {},
   "source": [
    "## 4. The test window\n",
    "\n",
    "Fit with the chosen settings on everything before the test window, then predict the test weeks."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "329d0777",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-11-10T05:50:27.411522Z",
     "iopub.status.busy": "2026-11-10T05:50:27.411266Z",
     "iopub.status.idle": "2026-11-10T05:50:27.569351Z",
     "shell.execute_reply": "2026-11-10T05:50:27.568698Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Error on the 12 unseen test weeks: 2.3%\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x320 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_all, scales = design(df, best_hyper)\n",
    "b0, beta = ridge_nonneg(X_all[: n - H], y_all[: n - H], len(channels), best_lam)\n",
    "pred = b0 + X_all @ beta\n",
    "test_mape = np.mean(np.abs((y_all[n - H:] - pred[n - H:]) / y_all[n - H:]))\n",
    "print(f\"Error on the {H} unseen test weeks: {test_mape:.1%}\")\n",
    "\n",
    "plt.figure(figsize=(10, 3.2))\n",
    "plt.plot(df[\"date\"], y_all, label=\"actual\", color=\"black\", lw=1)\n",
    "plt.plot(df[\"date\"], pred, label=\"model\", lw=2)\n",
    "plt.axvline(df[\"date\"].iloc[n - H], color=\"grey\", ls=\"--\", label=\"test window starts\")\n",
    "plt.title(\"Actual vs model\")\n",
    "plt.legend()\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "95c8737c",
   "metadata": {},
   "source": [
    "## 5. Final fit, decomposition, ROI and next-dollar return\n",
    "\n",
    "**ROI** is a channel's incremental revenue over the whole period divided by its spend. **Next-dollar return** is the slope of its response curve at recent weekly spend: what one more unit returns now. Budget decisions use the second."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "0cd5890d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-11-10T05:50:27.571712Z",
     "iopub.status.busy": "2026-11-10T05:50:27.571466Z",
     "iopub.status.idle": "2026-11-10T05:50:27.766419Z",
     "shell.execute_reply": "2026-11-10T05:50:27.765840Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Media drove 31% of revenue\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x320 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>contribution</th>\n",
       "      <th>roi</th>\n",
       "      <th>next_dollar</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>channel</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>google_search_spend</th>\n",
       "      <td>3918467.82</td>\n",
       "      <td>1.59</td>\n",
       "      <td>1.89</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>meta_spend</th>\n",
       "      <td>1495457.85</td>\n",
       "      <td>0.84</td>\n",
       "      <td>1.51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>offline_tv_spend</th>\n",
       "      <td>8059323.85</td>\n",
       "      <td>3.64</td>\n",
       "      <td>1.79</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     contribution   roi  next_dollar\n",
       "channel                                             \n",
       "google_search_spend    3918467.82  1.59         1.89\n",
       "meta_spend             1495457.85  0.84         1.51\n",
       "offline_tv_spend       8059323.85  3.64         1.79"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "b0, beta = ridge_nonneg(X_all, y_all, len(channels), best_lam)\n",
    "\n",
    "def steady_response(beta_j, hyper_j, scale_j, weekly):\n",
    "    decay, half, shape = hyper_j\n",
    "    return beta_j * hill(weekly / (1 - decay) / scale_j, half, shape)\n",
    "\n",
    "def summary(beta, hyper, scales):\n",
    "    rows = []\n",
    "    for j, ch in enumerate(channels):\n",
    "        contrib = beta[j] * X_all[:, j]\n",
    "        spend = df[ch].sum()\n",
    "        recent = df[ch].iloc[-13:].mean()\n",
    "        dx = recent * 0.01\n",
    "        m = (steady_response(beta[j], hyper[j], scales[j], recent + dx) - steady_response(beta[j], hyper[j], scales[j], recent)) / dx\n",
    "        rows.append(dict(channel=ch, contribution=contrib.sum(), roi=contrib.sum() / spend, next_dollar=m))\n",
    "    return pd.DataFrame(rows).set_index(\"channel\")\n",
    "\n",
    "point = summary(beta, best_hyper, scales)\n",
    "print(f\"Media drove {point.contribution.sum() / y_all.sum():.0%} of revenue\")\n",
    "\n",
    "fig, ax = plt.subplots(1, len(channels), figsize=(12, 3.2))\n",
    "for j, ch in enumerate(channels):\n",
    "    w = np.linspace(0, df[ch].max() * 2, 60)\n",
    "    ax[j].plot(w, steady_response(beta[j], best_hyper[j], scales[j], w))\n",
    "    ax[j].axvline(df[ch].iloc[-13:].mean(), color=\"black\", ls=\"--\", lw=1)\n",
    "    ax[j].axvline(df[ch].max(), color=\"orange\", ls=\"--\", lw=1)\n",
    "    ax[j].set(title=ch, xlabel=\"weekly spend\", ylabel=\"weekly revenue\")\n",
    "plt.suptitle(\"Response curves (black: recent spend, orange: highest week in the data)\")\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "point.round(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e2eafa07",
   "metadata": {},
   "source": [
    "## 6. Bootstrap ranges\n",
    "\n",
    "Refit on resampled data many times, and search the settings again each time, so the range includes the uncertainty in carryover and saturation, not just the coefficients. Residuals are resampled in 8-week blocks to keep their week-to-week pattern."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "d8f8f47a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-11-10T05:50:27.768849Z",
     "iopub.status.busy": "2026-11-10T05:50:27.768586Z",
     "iopub.status.idle": "2026-11-10T05:50:30.056090Z",
     "shell.execute_reply": "2026-11-10T05:50:30.055144Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">roi</th>\n",
       "      <th colspan=\"2\" halign=\"left\">next_dollar</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>0.05</th>\n",
       "      <th>0.95</th>\n",
       "      <th>0.05</th>\n",
       "      <th>0.95</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>channel</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>google_search_spend</th>\n",
       "      <td>1.15</td>\n",
       "      <td>2.41</td>\n",
       "      <td>1.42</td>\n",
       "      <td>2.23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>meta_spend</th>\n",
       "      <td>0.66</td>\n",
       "      <td>1.08</td>\n",
       "      <td>1.28</td>\n",
       "      <td>1.76</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>offline_tv_spend</th>\n",
       "      <td>2.61</td>\n",
       "      <td>5.59</td>\n",
       "      <td>1.37</td>\n",
       "      <td>2.47</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      roi       next_dollar      \n",
       "                     0.05  0.95        0.05  0.95\n",
       "channel                                          \n",
       "google_search_spend  1.15  2.41        1.42  2.23\n",
       "meta_spend           0.66  1.08        1.28  1.76\n",
       "offline_tv_spend     2.61  5.59        1.37  2.47"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "resid = y_all - (b0 + X_all @ beta)\n",
    "draws = []\n",
    "for b in range(40):\n",
    "    starts = rng.integers(0, n - 8, size=n // 8 + 1)\n",
    "    block = np.concatenate([resid[s:s + 8] for s in starts])[:n]\n",
    "    y_star = (b0 + X_all @ beta) + block\n",
    "    # small re-search around the best settings\n",
    "    best = None\n",
    "    for _ in range(8):\n",
    "        h = [(np.clip(d + rng.normal(0, 0.08), *BOUNDS[\"decay\"]), np.clip(hs + rng.normal(0, 0.1), *BOUNDS[\"half\"]), np.clip(sh + rng.normal(0, 0.2), *BOUNDS[\"shape\"])) for d, hs, sh in best_hyper]\n",
    "        s, lam = score(h, df.iloc[: n - H], y_star[: n - H])\n",
    "        if best is None or s < best[0]:\n",
    "            best = (s, lam, h)\n",
    "    Xb, scb = design(df, best[2])\n",
    "    _, bb = ridge_nonneg(Xb, y_star, len(channels), best[1])\n",
    "    X_all_saved = X_all\n",
    "    X_all = Xb\n",
    "    draws.append(summary(bb, best[2], scb))\n",
    "    X_all = X_all_saved\n",
    "\n",
    "boot = pd.concat(draws, keys=range(len(draws)))\n",
    "ranges = boot.groupby(level=1)[[\"roi\", \"next_dollar\"]].quantile([0.05, 0.95]).unstack().round(2)\n",
    "ranges"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "057f289a",
   "metadata": {},
   "source": [
    "## 7. Grade it against the answer key\n",
    "\n",
    "The synthetic data was simulated from these true ROIs (the lab's answer key shows the same). Does the 90% range contain them?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "d810b103",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-11-10T05:50:30.059931Z",
     "iopub.status.busy": "2026-11-10T05:50:30.059329Z",
     "iopub.status.idle": "2026-11-10T05:50:30.072661Z",
     "shell.execute_reply": "2026-11-10T05:50:30.072077Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>true_roi</th>\n",
       "      <th>model_roi</th>\n",
       "      <th>low</th>\n",
       "      <th>high</th>\n",
       "      <th>contains_truth</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>google_search_spend</th>\n",
       "      <td>1.98</td>\n",
       "      <td>1.59</td>\n",
       "      <td>1.15</td>\n",
       "      <td>2.41</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>meta_spend</th>\n",
       "      <td>1.05</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.66</td>\n",
       "      <td>1.08</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>offline_tv_spend</th>\n",
       "      <td>1.63</td>\n",
       "      <td>3.64</td>\n",
       "      <td>2.61</td>\n",
       "      <td>5.59</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     true_roi  model_roi   low  high  contains_truth\n",
       "google_search_spend      1.98       1.59  1.15  2.41            True\n",
       "meta_spend               1.05       0.84  0.66  1.08            True\n",
       "offline_tv_spend         1.63       3.64  2.61  5.59           False"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "true_roi = {\"google_search_spend\": 1.98, \"meta_spend\": 1.05, \"offline_tv_spend\": 1.63}\n",
    "check = pd.DataFrame({\n",
    "    \"true_roi\": true_roi,\n",
    "    \"model_roi\": point.roi,\n",
    "    \"low\": ranges[(\"roi\", 0.05)],\n",
    "    \"high\": ranges[(\"roi\", 0.95)],\n",
    "})\n",
    "check[\"contains_truth\"] = (check.true_roi >= check.low) & (check.true_roi <= check.high)\n",
    "check.round(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3856b042",
   "metadata": {},
   "source": [
    "If one channel's range misses the truth, you have just seen why the lab pools four independent searches: a single search can land on a different local optimum when the data only weakly shows saturation. Try a few seeds and watch which channel moves."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "987e7d20",
   "metadata": {},
   "source": [
    "## What to try next\n",
    "\n",
    "* Change `rng = np.random.default_rng(42)` to another seed and rerun. How much do the settings and ROIs move? (The lab pools four seeds to make its answer stable.)\n",
    "* Delete `promo` from `controls`. Which channel absorbs the promotion effect?\n",
    "* Make Google and Meta spend move together (for example, copy one into the other with noise). Watch the ranges widen: that is collinearity.\n",
    "* Compare with the lab's numbers on the same CSV at [realanalystjobs.com/learn/marketing-mix-modeling](https://realanalystjobs.com/learn/marketing-mix-modeling)."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.9.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}
