Skip to content

Instantly share code, notes, and snippets.

@mrocklin

mrocklin/deltalake-play.ipynb

Created February 8, 2024 14:53
Show Gist options

  • Select an option

    Learn more about clone URLs
  • Save mrocklin/c97bdc584c8568dc07c48b2c38c00fd2 to your computer and use it in GitHub Desktop.

Select an option

Learn more about clone URLs
Save mrocklin/c97bdc584c8568dc07c48b2c38c00fd2 to your computer and use it in GitHub Desktop.
Download ZIP
Display the source blob
Display the rendered blob
Raw
deltalake-play.ipynb
{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"id": "6caaa3f8-b713-4188-b237-557d071fa14d",
"metadata": {},
"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>name</th>\n",
" <th>id</th>\n",
" <th>x</th>\n",
" <th>y</th>\n",
" </tr>\n",
" <tr>\n",
" <th>timestamp</th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" <th></th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>2000-01-01 00:00:00</th>\n",
" <td>Frank</td>\n",
" <td>998</td>\n",
" <td>-0.879896</td>\n",
" <td>0.438060</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2000-01-01 00:00:01</th>\n",
" <td>Victor</td>\n",
" <td>1015</td>\n",
" <td>0.246495</td>\n",
" <td>-0.190725</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2000-01-01 00:00:02</th>\n",
" <td>Charlie</td>\n",
" <td>1008</td>\n",
" <td>0.173829</td>\n",
" <td>0.278618</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2000-01-01 00:00:03</th>\n",
" <td>Victor</td>\n",
" <td>997</td>\n",
" <td>0.668464</td>\n",
" <td>0.126580</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2000-01-01 00:00:04</th>\n",
" <td>Charlie</td>\n",
" <td>1008</td>\n",
" <td>0.776633</td>\n",
" <td>-0.318526</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" name id x y\n",
"timestamp \n",
"2000-01-01 00:00:00 Frank 998 -0.879896 0.438060\n",
"2000-01-01 00:00:01 Victor 1015 0.246495 -0.190725\n",
"2000-01-01 00:00:02 Charlie 1008 0.173829 0.278618\n",
"2000-01-01 00:00:03 Victor 997 0.668464 0.126580\n",
"2000-01-01 00:00:04 Charlie 1008 0.776633 -0.318526"
]
},
"execution_count": 1,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"import deltalake, dask\n",
"df = dask.datasets.timeseries().compute()\n",
"df.head()"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "5558be08-ec55-460e-8bf1-8fc2b778c969",
"metadata": {},
"outputs": [],
"source": [
"!rm -rf tmp"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "8edab4d4-d321-4b80-a3a1-41a2fa73f36c",
"metadata": {},
"outputs": [],
"source": [
"deltalake.write_deltalake(\"tmp\", df, mode=\"append\")"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "4c0df6a5-f728-4a0d-b56e-393a2a80db80",
"metadata": {},
"outputs": [],
"source": [
"deltalake.write_deltalake(\"tmp\", df, mode=\"append\")"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "bbc6f3ca-8f63-49cb-9494-70d79ab42016",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"DeltaTable()"
]
},
"execution_count": 5,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"t = deltalake.DeltaTable(\"tmp\")\n",
"t"
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "aa090e17-ab92-46c3-b5d5-5ee761a0667e",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"['1-a5fab5c0-47d6-4603-adcd-9b310a0d36c7-0.parquet',\n",
" '0-dcfe9ff9-8046-4dd0-b285-32ec95539df7-0.parquet']"
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"t.files()"
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "849e4bfc-7094-4bdb-84d1-4ff7cdfb5dc2",
"metadata": {},
"outputs": [],
"source": [
"for _ in range(10):\n",
" deltalake.write_deltalake(\"tmp\", df.head(), mode=\"append\")"
]
},
{
"cell_type": "code",
"execution_count": 8,
"id": "84b7b53b-6217-4e6e-b1d1-20681c56fd6b",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"['1-a5fab5c0-47d6-4603-adcd-9b310a0d36c7-0.parquet',\n",
" '0-dcfe9ff9-8046-4dd0-b285-32ec95539df7-0.parquet']"
]
},
"execution_count": 8,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"t.files()"
]
},
{
"cell_type": "code",
"execution_count": 9,
"id": "e8c5f4e0-4d69-4965-beee-0069e24e9c8c",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"['11-7acbd002-7713-4309-89a4-7c6e4fd762d8-0.parquet',\n",
" '10-39e28cdb-004c-4de0-8cbe-da52993e3688-0.parquet',\n",
" '9-77fe16e2-4d5d-4c62-89be-a6f89856722b-0.parquet',\n",
" '8-5da355e8-86cc-417a-9f65-e3ab11d3dde0-0.parquet',\n",
" '7-59ff1cb9-8ea6-424a-ba4b-846e329a27ce-0.parquet',\n",
" '6-8982c38b-b51b-43f2-b839-bc06e11df490-0.parquet',\n",
" '5-aab946db-1ba5-4b67-9c10-41a984109f0e-0.parquet',\n",
" '4-725e0842-41f2-4172-b96a-aa7c690cf3a0-0.parquet',\n",
" '3-d2cf83b5-33db-4c65-af31-f65daa2d1851-0.parquet',\n",
" '2-33bcaf9c-b3f6-4a88-8eba-ca0cd2489ea4-0.parquet',\n",
" '1-a5fab5c0-47d6-4603-adcd-9b310a0d36c7-0.parquet',\n",
" '0-dcfe9ff9-8046-4dd0-b285-32ec95539df7-0.parquet']"
]
},
"execution_count": 9,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"t = deltalake.DeltaTable(\"tmp\")\n",
"t.files()"
]
},
{
"cell_type": "code",
"execution_count": 10,
"id": "2ffb8e1d-8825-44b0-977c-188570431811",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"{'numFilesAdded': 1,\n",
" 'numFilesRemoved': 11,\n",
" 'filesAdded': {'min': 53376152,\n",
" 'max': 53376152,\n",
" 'avg': 53376152.0,\n",
" 'totalFiles': 1,\n",
" 'totalSize': 53376152},\n",
" 'filesRemoved': {'min': 2118,\n",
" 'max': 77924229,\n",
" 'avg': 7085946.2727272725,\n",
" 'totalFiles': 11,\n",
" 'totalSize': 77945409},\n",
" 'partitionsOptimized': 1,\n",
" 'numBatches': 2558,\n",
" 'totalConsideredFiles': 12,\n",
" 'totalFilesSkipped': 1,\n",
" 'preserveInsertionOrder': True}"
]
},
"execution_count": 10,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"t.optimize.compact()"
]
},
{
"cell_type": "code",
"execution_count": 11,
"id": "9da09aba-1c71-4bd6-870f-05c3b93cd1c9",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"['part-00001-a246e865-bb70-4367-89e8-533773f32891-c000.zstd.parquet',\n",
" '0-dcfe9ff9-8046-4dd0-b285-32ec95539df7-0.parquet']"
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"t.files()"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python [conda env:etl]",
"language": "python",
"name": "conda-env-etl-py"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.12.1"
}
},
"nbformat": 4,
"nbformat_minor": 5
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment