dataforcanada/d4c-datapkg-statistical
1
0
Fork 0
mirror of https://github.com/dataforcanada/d4c-datapkg-statistical.git synced 2026-06-13 14:10:55 +02:00
Code Issues Packages Projects Releases Wiki Activity

Format DuckDB SQL code

Browse Source
This commit is contained in:
Diego Ripley
2025-05-25 20:05:15 +00:00
parent b4b7e20219
commit d4a25af7b1
1 changed files with 35 additions and 21 deletions
Download Patch File Download Diff File Expand all files Collapse all files
experiments/duckdb_census_of_population.ipynb
+34 -20
View File
@@ -2,28 +2,19 @@
"cells": [
{
"cell_type": "code",
"execution_count": 36,
"execution_count": 40,
"id": "56ac906e",
"metadata": {},
"outputs": [],
"source": [
"import buckaroo\n",
"import duckdb\n",
"from IPython.core.interactiveshell import InteractiveShell\n",
"from lonboard import viz\n",
"import pandas as pd\n",
"\n",
"# Enable multiple outputs per cell\n",
"InteractiveShell.ast_node_interactivity = \"all\"\n",
"# Show all columns\n",
"pd.set_option('display.max_columns', None)\n",
"\n",
"output_data_folder = '/data/experiments'"
]
},
{
"cell_type": "code",
"execution_count": 2,
"execution_count": 41,
"id": "708e293d",
"metadata": {},
"outputs": [],
@@ -44,17 +35,31 @@
},
{
"cell_type": "code",
"execution_count": 28,
"execution_count": 42,
"id": "580c82ad-f64d-439f-9055-2307fdf7cccd",
"metadata": {},
"outputs": [
{
"data": {
"application/vnd.jupyter.widget-view+json": {
"model_id": "8c53af935cf44664a8b493f5133b87e7",
"version_major": 2,
"version_minor": 0
},
"text/plain": [
"<duckdb.duckdb.DuckDBPyConnection at 0x7f0704bb5e70>"
"FloatProgress(value=0.0, layout=Layout(width='auto'), style=ProgressStyle(bar_color='black'))"
]
},
"execution_count": 28,
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/plain": [
"<duckdb.duckdb.DuckDBPyConnection at 0x7f0654aa5730>"
]
},
"execution_count": 42,
"metadata": {},
"output_type": "execute_result"
}
@@ -63,10 +68,19 @@
"con.execute(\"\"\"\n",
"DROP TABLE IF EXISTS geo_data;\n",
"CREATE TABLE geo_data AS\n",
"SELECT geo.da_dguid, cop.count_total_4, CAST(round((cop.count_total_1 / (ST_Area_Spheroid(geo.geom) / 1000000.0)), 0) AS INTEGER) AS count_total_4_per_square_km, geo.geom \n",
"FROM 'https://data.dataforcanada.org/processed/statistics_canada/census_of_population/2021/tabular/da_2021.parquet' AS cop,\n",
"SELECT\n",
" geo.da_dguid,\n",
" cop.count_total_4,\n",
" CAST(\n",
" ROUND(\n",
" (cop.count_total_1 / (ST_AREA_SPHEROID(geo.geom) / 1000000.0)), 0\n",
" ) AS INTEGER\n",
" ) AS count_total_4_per_square_km,\n",
" geo.geom\n",
"FROM\n",
" 'https://data.dataforcanada.org/processed/statistics_canada/census_of_population/2021/tabular/da_2021.parquet' AS cop,\n",
" 'https://data.dataforcanada.org/processed/statistics_canada/boundaries/2021/digital_boundary_files/da_2021.parquet' AS geo\n",
"WHERE geo.csd_name in ('Ottawa') AND cop.da_dguid = geo.da_dguid;\n",
"WHERE geo.csd_name IN ('Ottawa') AND cop.da_dguid = geo.da_dguid;\n",
"\"\"\")"
]
},
@@ -80,17 +94,17 @@
},
{
"cell_type": "code",
"execution_count": 37,
"execution_count": 43,
"id": "0528eeb4-fb89-4640-94c6-c0eda4cbe764",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"<duckdb.duckdb.DuckDBPyConnection at 0x7f0704bb5e70>"
"<duckdb.duckdb.DuckDBPyConnection at 0x7f0654aa5730>"
]
},
"execution_count": 37,
"execution_count": 43,
"metadata": {},
"output_type": "execute_result"
}