{
"cells": [
{
"cell_type": "code",
"execution_count": 2,
"id": "initial_id",
"metadata": {
"ExecuteTime": {
"end_time": "2025-07-08T15:58:03.741302Z",
"start_time": "2025-07-08T15:58:03.737218Z"
},
"collapsed": true
},
"outputs": [],
"source": [
"import pandas as pd\n",
"import matplotlib.pyplot as plt\n",
"import seaborn as sns\n",
"import sys\n",
"import os\n",
"import importlib\n",
"\n",
"sys.path.append(os.path.abspath(\"..\"))\n",
"import misc\n",
"importlib.reload(misc)\n",
"from misc import load_csv_dataset"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "74a8a262bc55d976",
"metadata": {
"ExecuteTime": {
"end_time": "2025-07-08T15:58:27.047640Z",
"start_time": "2025-07-08T15:58:06.842257Z"
}
},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
">> Loading CSV dataset from names_featured.csv\n",
">> Detected fieldnames: ['name', 'sex', 'region', 'year', 'words', 'length', 'probable_native', 'probable_surname', 'identified_category', 'identified_name', 'identified_surname', 'annotated', 'province']\n",
">> Successfully loaded with UTF-8 encoding\n"
]
}
],
"source": [
"df = pd.DataFrame(load_csv_dataset(\"names_featured.csv\"))"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "ebb18cabd40b011f",
"metadata": {
"ExecuteTime": {
"end_time": "2025-07-08T15:58:50.742471Z",
"start_time": "2025-07-08T15:58:45.080832Z"
}
},
"outputs": [],
"source": [
"df['year'] = pd.to_numeric(df['year'], errors='coerce')\n",
"df['words'] = pd.to_numeric(df['words'], errors='coerce')\n",
"df['length'] = pd.to_numeric(df['length'], errors='coerce')"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "295eda175081129c",
"metadata": {
"ExecuteTime": {
"end_time": "2025-07-08T15:59:02.783458Z",
"start_time": "2025-07-08T15:59:02.297971Z"
}
},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" | \n",
" count | \n",
" mean | \n",
" std | \n",
" min | \n",
" 25% | \n",
" 50% | \n",
" 75% | \n",
" max | \n",
"
\n",
" \n",
" \n",
" \n",
" | year | \n",
" 5522918.0 | \n",
" 2017.037288 | \n",
" 4.657671 | \n",
" 2008.0 | \n",
" 2013.0 | \n",
" 2018.0 | \n",
" 2021.0 | \n",
" 2023.0 | \n",
"
\n",
" \n",
" | words | \n",
" 5522918.0 | \n",
" 2.862474 | \n",
" 0.473505 | \n",
" 1.0 | \n",
" 3.0 | \n",
" 3.0 | \n",
" 3.0 | \n",
" 10.0 | \n",
"
\n",
" \n",
" | length | \n",
" 5522918.0 | \n",
" 18.217497 | \n",
" 3.498697 | \n",
" 1.0 | \n",
" 16.0 | \n",
" 18.0 | \n",
" 21.0 | \n",
" 49.0 | \n",
"
\n",
" \n",
"
\n",
"