299 lines
12 KiB
Plaintext
299 lines
12 KiB
Plaintext
{
|
|
"cells": [
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 1,
|
|
"id": "e9bf1d26",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"from pyspark.sql import SparkSession, DataFrame\n",
|
|
"from pyspark.sql import functions as F\n",
|
|
"from pathlib import Path\n",
|
|
"import glob\n",
|
|
"import os\n",
|
|
"from dotenv import load_dotenv, find_dotenv"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 2,
|
|
"id": "11bb5dc2",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
" ## Extract ##\n",
|
|
"\n",
|
|
"BASE = Path.cwd() / \"ingest\"\n",
|
|
"\n",
|
|
"def start_spark(app_name):\n",
|
|
" jar_path = str(Path.cwd() / \"postgresql.jar\")\n",
|
|
" spark = SparkSession.builder.appName(app_name).config(\"spark.jars\", jar_path).config(\"spark.driver.extraClassPath\", jar_path).getOrCreate()\n",
|
|
" spark.sparkContext.setLogLevel(\"ERROR\")\n",
|
|
" return spark\n",
|
|
"\n",
|
|
"def music_source_paths():\n",
|
|
" return sorted(glob.glob(str(BASE / \"*_sesh\" / \"*.json\")))\n",
|
|
"\n",
|
|
"def video_source_paths():\n",
|
|
" return sorted(glob.glob(str(BASE / \"trakt_*\" / \"*.json\")))\n",
|
|
"\n",
|
|
"def read_json(spark, paths):\n",
|
|
" if not paths:\n",
|
|
" raise FileNotFoundError(\"No JSON files matched the pattern\")\n",
|
|
" return (\n",
|
|
" spark.read\n",
|
|
" .format(\"json\")\n",
|
|
" .option(\"multiLine\", True)\n",
|
|
" .json(paths) # list of paths, one read\n",
|
|
" # .withColumn(\"source_file\", F.input_file_name())\n",
|
|
" # .withColumn(\n",
|
|
" # \"source_folder\",\n",
|
|
" # F.regexp_extract(\"source_file\", r\"/ingest/([^/]+)/\", 1),\n",
|
|
" # )\n",
|
|
" )"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 3,
|
|
"id": "a898326e",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
" ## Transform ##\n",
|
|
"def null_counts_by_column(df):\n",
|
|
" null_counts = df.select([\n",
|
|
" F.count(F.when(F.col(c).isNull(), c)).alias(c)\n",
|
|
" for c in df.columns\n",
|
|
" ])\n",
|
|
" return null_counts\n",
|
|
"\n",
|
|
"def metadata_df(df):\n",
|
|
" return df.withColumnRenamed(\"master_metadata_album_album_name\", \"album_name\").withColumnRenamed(\"master_metadata_album_artist_name\", \"artist_name\").withColumnRenamed(\"master_metadata_track_name\", \"track_name\")\n",
|
|
"\n",
|
|
"def rename_columns(df, rename_map, drop_columns):\n",
|
|
" missing_rename = set(rename_map) - set(df.columns)\n",
|
|
" missing_dropped = set(drop_columns) - set(df.columns)\n",
|
|
"\n",
|
|
" if missing_rename:\n",
|
|
" print(f\"Warning: columns not found: {missing_rename}\")\n",
|
|
" if missing_dropped:\n",
|
|
" print(f\"Warning: columns not found: {missing_dropped}\")\n",
|
|
"\n",
|
|
" df = df.drop(*drop_columns)\n",
|
|
" \n",
|
|
" return df.toDF(*[rename_map.get(c, c) for c in df.columns])\n",
|
|
"\n",
|
|
"def track_df(df):\n",
|
|
" return df.where(df[\"track_name\"].isNotNull())\n",
|
|
"\n",
|
|
"def parse_timestamps(df: DataFrame, columns) -> DataFrame:\n",
|
|
" for column in columns:\n",
|
|
" df = df.withColumn(column, F.to_timestamp(column))\n",
|
|
" return df\n",
|
|
"\n",
|
|
"def parse_booleans(df: DataFrame, columns) -> DataFrame:\n",
|
|
" df = df.select([F.col(c).cast(\"string\") for c in columns])\n",
|
|
" return df\n",
|
|
"\n",
|
|
"def add_minutes_played(df: DataFrame, src, dest) -> DataFrame:\n",
|
|
" return (df.withColumn(dest, F.round(F.col(src) / 60000, 2)))\n",
|
|
"\n",
|
|
"def add_time_parts(df: DataFrame, ts_col) -> DataFrame:\n",
|
|
" return (df\n",
|
|
" .withColumn(\"ts\", F.to_timestamp(\"ts\"))\n",
|
|
" .withColumn(\"offline_timestamped\", F.to_timestamp(\"offline_timestamp\"))\n",
|
|
" .withColumn('time', F.date_format('ts', 'HH:mm:ss'))\n",
|
|
" .withColumn('date', F.date_format('ts', 'yyyy-MM-dd'))\n",
|
|
" .withColumn('offline_time', F.date_format('offline_timestamped', 'HH:mm:ss'))\n",
|
|
" .withColumn('offline_date', F.date_format('offline_timestamped', 'yyyy-MM-dd')) \n",
|
|
" )\n",
|
|
"\n",
|
|
"def clean_track_df(df: DataFrame) -> DataFrame:\n",
|
|
" return(df\n",
|
|
" .transform(rename_columns, SPOTIFY_RENAMES, MUSIC_DROP_COLUMNS)\n",
|
|
" .transform(track_df)\n",
|
|
" # .transform(parse_timestamps, [\"ts\", \"offline_timestamp\"])\n",
|
|
" # .transform(parse_booleans, [\"offline\", \"shuffle\", \"skipped\"]) ##returns only the selected columns\n",
|
|
" .transform(add_minutes_played, \"ms_played\", \"ms_played\")\n",
|
|
" .transform(add_time_parts, \"ts\")\n",
|
|
" )"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 4,
|
|
"id": "817475cf",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
" ## Load ##\n",
|
|
"def get_jdbc_config():\n",
|
|
" \"\"\"Load env vars and return a Postgres URL\"\"\"\n",
|
|
" load_dotenv(find_dotenv())\n",
|
|
" \n",
|
|
" username=os.getenv(\"DATABASE_USERNAME\")\n",
|
|
" password=os.getenv(\"DATABASE_PASSWORD\")\n",
|
|
" host=os.getenv(\"DATABASE_IP\")\n",
|
|
" port=int(os.getenv(\"DATABASE_PORT\", 5432))\n",
|
|
" database=os.getenv(\"DATABASE_NAME\")\n",
|
|
" \n",
|
|
" missing = [k for k, v in {\n",
|
|
" \"DATABASE_USERNAME\": username,\n",
|
|
" \"DATABASE_PASSWORD\": password,\n",
|
|
" \"DATABASE_IP\": host,\n",
|
|
" \"DATABASE_NAME\": database,}.items() if not v]\n",
|
|
" if missing:\n",
|
|
" raise ValueError(f\"Missing env vars: {missing}\")\n",
|
|
"\n",
|
|
" url = f\"jdbc:postgresql://{host}:{port}/{database}\"\n",
|
|
" properties = {\n",
|
|
" \"user\": username,\n",
|
|
" \"password\": password,\n",
|
|
" \"driver\": \"org.postgresql.Driver\"\n",
|
|
" }\n",
|
|
" return url, properties\n",
|
|
"\n",
|
|
"def load_data(df, table_name):\n",
|
|
" \"\"\"Load the data into the database.\"\"\"\n",
|
|
" url, properties = get_jdbc_config()\n",
|
|
" df.write.jdbc(url=url, table=table_name, mode=\"overwrite\", properties=properties)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 5,
|
|
"id": "c609291c",
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"## Config \n",
|
|
"SPOTIFY_RENAMES = {\n",
|
|
" \"master_metadata_album_album_name\": \"album_name\",\n",
|
|
" \"master_metadata_album_artist_name\": \"artist_name\",\n",
|
|
" \"master_metadata_track_name\": \"track_name\",\n",
|
|
" \"spotify_episode_uri\" : \"episode_uri\",\n",
|
|
" \"spotify_track_uri\" : \"track_uri\"\n",
|
|
"}\n",
|
|
"\n",
|
|
"TRAKT_RENAMES = {\"watched_at\": \"ts\", \"title\": \"video_title\"}\n",
|
|
"\n",
|
|
"MUSIC_DROP_COLUMNS = [\"ip_addr_decrypted\", \"user_agent_decrypted\", \"incognito_mode\", \"conn_country\", \"spotify_episode_uri\", \"episode_name\", \"episode_show_name\"]\n",
|
|
"\n"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"id": "4cc14bd0",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"WARNING: Using incubator modules: jdk.incubator.vector\n",
|
|
"Using Spark's default log4j profile: org/apache/spark/log4j2-defaults.properties\n",
|
|
"26/10/18 17:57:09 WARN Utils: Your hostname, MacBook-Neo-232.local, resolves to a loopback address: 127.0.0.1; using 192.168.10.130 instead (on interface en0)\n",
|
|
"26/10/18 17:57:09 WARN Utils: Set SPARK_LOCAL_IP if you need to bind to another address\n",
|
|
"26/10/18 17:57:09 WARN NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable\n",
|
|
"Using Spark's default log4j profile: org/apache/spark/log4j2-defaults.properties\n",
|
|
"Setting default log level to \"WARN\".\n",
|
|
"To adjust logging level use sc.setLogLevel(newLevel). For SparkR, use setLogLevel(newLevel).\n",
|
|
" \r"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
" Raw DF Rows: 105477\n",
|
|
" Transformed DF Rows: 102626\n",
|
|
" Row difference is 2851\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
" ## MAIN ## \n",
|
|
"def main():\n",
|
|
" app_name = \"ml_media\"\n",
|
|
" spark = start_spark(app_name)\n",
|
|
"\n",
|
|
" table_name = \"spotify_data\"\n",
|
|
" raw_df = read_json(spark, music_source_paths())\n",
|
|
" transformed_df = clean_track_df(raw_df)\n",
|
|
" load_data(transformed_df, table_name)\n",
|
|
"\n",
|
|
" print(f\" Raw DF Rows: {raw_df.count()}\")\n",
|
|
" print(f\" Transformed DF Rows: {transformed_df.count()}\")\n",
|
|
" print(f\" Row difference is {raw_df.count() - transformed_df.count()}\") \n",
|
|
" spark.stop()\n",
|
|
"if __name__ == \"__main__\":\n",
|
|
" main()\n",
|
|
"\n",
|
|
"\n",
|
|
"# transformed_df.groupBy(\"username\").count().show()"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 7,
|
|
"id": "faa6a8cb",
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"ename": "NameError",
|
|
"evalue": "name 'music_df' is not defined",
|
|
"output_type": "error",
|
|
"traceback": [
|
|
"\u001b[31m---------------------------------------------------------------------------\u001b[39m",
|
|
"\u001b[31mNameError\u001b[39m Traceback (most recent call last)",
|
|
"\u001b[36mCell\u001b[39m\u001b[36m \u001b[39m\u001b[32mIn[7]\u001b[39m\u001b[32m, line 2\u001b[39m\n\u001b[32m 1\u001b[39m \u001b[38;5;66;03m## Potential Test Help\u001b[39;00m\n\u001b[32m----> \u001b[39m\u001b[32m2\u001b[39m print(metadata_df(music_df).rdd.getNumPartitions()) \u001b[38;5;66;03m# Partitions for Pyspark\u001b[39;00m\n\u001b[32m 3\u001b[39m print(len(metadata_df(music_df).columns)) \u001b[38;5;66;03m# # of columns in JSON\u001b[39;00m\n\u001b[32m 4\u001b[39m null_counts = null_counts_by_column(metadata_df(music_df)) \u001b[38;5;66;03m# Null counts by column\u001b[39;00m\n\u001b[32m 5\u001b[39m null_counts.show(vertical=\u001b[38;5;28;01mTrue\u001b[39;00m) \u001b[38;5;66;03m# Print Null counts by column_df\u001b[39;00m\n",
|
|
"\u001b[31mNameError\u001b[39m: name 'music_df' is not defined"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"## Potential Test Help\n",
|
|
"print(metadata_df(music_df).rdd.getNumPartitions()) # Partitions for Pyspark\n",
|
|
"print(len(metadata_df(music_df).columns)) # # of columns in JSON\n",
|
|
"null_counts = null_counts_by_column(metadata_df(music_df)) # Null counts by column\n",
|
|
"null_counts.show(vertical=True) # Print Null counts by column_df\n",
|
|
"total_rows = music_df.count\n",
|
|
"\n",
|
|
"df.count()-track_df.count() ## assert the difference in rows for tracks is equal to the nulls count\n",
|
|
" ## same can be done for podcasts\n",
|
|
"\n",
|
|
"df.show(5)\n",
|
|
"print(f\"Original Rows {df.count()}\")\n",
|
|
"\n",
|
|
"track_df.show(5)\n",
|
|
"print(f\"Track Rows {track_df.count()}\")\n",
|
|
"\n"
|
|
]
|
|
}
|
|
],
|
|
"metadata": {
|
|
"kernelspec": {
|
|
"display_name": ".venv",
|
|
"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": "3.13.7"
|
|
}
|
|
},
|
|
"nbformat": 4,
|
|
"nbformat_minor": 5
|
|
}
|