{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Task 0 — Datasaurus Warm‑Up (Starter)\n",
    "\n",
    "**Goal:** Show why we must *always* visualize by comparing groups with nearly identical summary statistics but very different shapes when plotted.\n",
    "\n",
    "This starter uses `datasaurus_task0.csv` (long format) with four groups: `dino`, `star`, `circle`, `bullseye`.\n",
    "\n",
    "**What to do:**\n",
    "1. Load the CSV (as strings first).\n",
    "2. Compute basic stats per group (mean, std, correlation).\n",
    "3. Generate a SweetViz report (optional but recommended).\n",
    "4. Plot x vs y for each group (facet grid).\n",
    "5. Write *2–4 sentences* reflecting on why summary stats were misleading.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 0) (Optional) Install packages in this environment\n",
    "Uncomment if needed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "tags": [
     "setup"
    ]
   },
   "outputs": [],
   "source": [
    "# !pip install -q numpy pandas seaborn matplotlib sweetviz dtale\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1) Load the dataset\n",
    "Load as strings first (safer), then coerce numeric columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {},
   "outputs": [
    {
     "ename": "",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31mRunning cells with '.venv (Python 3.11.9)' requires the ipykernel package.\n",
      "\u001b[1;31mInstall 'ipykernel' into the Python environment. \n",
      "\u001b[1;31mCommand: 'd:/Projects/43679_InteractiveVis/VI_Lab_01_EDA/.venv/Scripts/python.exe -m pip install ipykernel -U --force-reinstall'"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "csv_path = 'datasaurus_task0.csv'\n",
    "df_raw = pd.read_csv(csv_path, dtype=str)\n",
    "df_raw.head()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Coerce numeric columns and quick sanity checks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df_raw.copy()\n",
    "for c in ['x','y']:\n",
    "    df[c] = pd.to_numeric(df[c], errors='coerce')\n",
    "df.info()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2) Basic summary stats by group (fill in)\n",
    "Compute mean, std for x & y by `dataset`, and the correlation within each group."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {},
   "outputs": [],
   "source": [
    "# TODO: groupby summaries\n",
    "# g = df.groupby('dataset')\n",
    "# means = g[['x','y']].mean()\n",
    "# stds  = g[['x','y']].std()\n",
    "# corr  = g.apply(lambda d: d[['x','y']].corr().iloc[0,1])\n",
    "# means, stds, corr\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3) (Optional) SweetViz profile\n",
    "Generate a quick report to observe that top-level stats look very similar."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import sweetviz as sv\n",
    "# report = sv.analyze(df)\n",
    "# report.show_html('task0_sweetviz_report.html')\n",
    "# print('Wrote task0_sweetviz_report.html')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4) Visualize — scatter by group (facet)\n",
    "Create a facet grid with one subplot per dataset and compare shapes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "sns.set_theme(style='white', context='notebook')\n",
    "g = sns.FacetGrid(df, col='dataset', col_wrap=2, height=4, sharex=True, sharey=True)\n",
    "g.map_dataframe(sns.scatterplot, x='x', y='y', s=20, edgecolor=None)\n",
    "g.set_titles('{col_name}')\n",
    "for ax in g.axes.flatten():\n",
    "    ax.set_xlabel('x'); ax.set_ylabel('y')\n",
    "plt.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5) Reflection (write here)\n",
    "**Prompt:** If the per-group mean/variance/correlation were similar, why do the plots look different?\n",
    "- Which shapes do you observe?\n",
    "- What does this imply for relying solely on `.describe()` or correlation before plotting?\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": ".venv",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "name": "python",
   "version": "3.11.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}