CC/media_ml
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

local vscode

Browse Source
This commit is contained in:
CC
2026-10-18 22:25:12 +01:00
parent f409cefa18
commit 7b6803b1ee
2 changed files with 51 additions and 56 deletions
Download Patch File Download Diff File Expand all files Collapse all files

65
spotify_bronze_silver.ipynb
View File

@@ -187,7 +187,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": 91,
"id": "4cc14bd0", "id": "4cc14bd0",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
@@ -201,46 +201,21 @@
"| GB| NULL| NULL| false| 31.94.22.158| Last Day Of Summer| Summer Walker| Girls Need Love| 80505| false| 1680818230| ios|unexpected-exit-w...| trackdone| true| false| NULL|spotify:track:644...|2023-04-06T21:58:31Z| unknown| csm3-us|\n", "| GB| NULL| NULL| false| 31.94.22.158| Last Day Of Summer| Summer Walker| Girls Need Love| 80505| false| 1680818230| ios|unexpected-exit-w...| trackdone| true| false| NULL|spotify:track:644...|2023-04-06T21:58:31Z| unknown| csm3-us|\n",
"+------------+------------+-----------------+--------------+-----------------+--------------------------------+---------------------------------+--------------------------+---------+-------+-----------------+--------+--------------------+------------+-------+-------+-------------------+--------------------+--------------------+--------------------+--------+\n", "+------------+------------+-----------------+--------------+-----------------+--------------------------------+---------------------------------+--------------------------+---------+-------+-----------------+--------+--------------------+------------+-------+-------+-------------------+--------------------+--------------------+--------------------+--------+\n",
"only showing top 1 row\n", "only showing top 1 row\n",
"+-------+--------------------+-----+\n", "+----------------------------+--------------+\n",
"|skipped| reason_end|count|\n", "|reason_end |skipped_values|\n",
"+-------+--------------------+-----+\n", "+----------------------------+--------------+\n",
"| false| unknown| 99|\n", "|NULL |[] |\n",
"| NULL| unknown| 375|\n", "|backbtn |[true] |\n",
"| false|unexpected-exit-w...| 1243|\n", "|endplay |[true] |\n",
"| NULL|unexpected-exit-w...| 432|\n", "|fwdbtn |[true] |\n",
"| false| unexpected-exit| 36|\n", "|logout |[false] |\n",
"| NULL| unexpected-exit| 51|\n", "|remote |[false] |\n",
"| false| trackerror| 69|\n", "|trackdone |[false] |\n",
"| NULL| trackerror| 33|\n", "|trackerror |[false] |\n",
"| false| trackdone|36102|\n", "|unexpected-exit |[false] |\n",
"| NULL| trackdone|24077|\n", "|unexpected-exit-while-paused|[false] |\n",
"| false| remote| 121|\n", "|unknown |[false] |\n",
"| NULL| remote| 859|\n", "+----------------------------+--------------+\n",
"| false| logout| 491|\n",
"| NULL| logout| 745|\n",
"| true| fwdbtn|16571|\n",
"| NULL| fwdbtn|11445|\n",
"| true| endplay| 2943|\n",
"| NULL| endplay| 1447|\n",
"| true| backbtn| 650|\n",
"| NULL| backbtn| 402|\n",
"+-------+--------------------+-----+\n",
"only showing top 20 rows\n",
"+----------------------------+-----+\n",
"|reason_end |count|\n",
"+----------------------------+-----+\n",
"|trackdone |60179|\n",
"|fwdbtn |28016|\n",
"|NULL |4435 |\n",
"|endplay |4390 |\n",
"|unexpected-exit-while-paused|1675 |\n",
"|logout |1236 |\n",
"|backbtn |1052 |\n",
"|remote |980 |\n",
"|unknown |474 |\n",
"|trackerror |102 |\n",
"|unexpected-exit |87 |\n",
"+----------------------------+-----+\n",
"\n", "\n",
" Raw DF Rows: 105477\n", " Raw DF Rows: 105477\n",
" Transformed DF Rows: 102626\n", " Transformed DF Rows: 102626\n",
@@ -260,10 +235,14 @@
" transformed_df = clean_track_df(raw_df)\n", " transformed_df = clean_track_df(raw_df)\n",
" #load_data(transformed_df, table_name)\n", " #load_data(transformed_df, table_name)\n",
" # transformed_df.show(5)\n", " # transformed_df.show(5)\n",
" skip_reason_df = transformed_df.groupBy(\"skipped\", \"reason_end\").count().orderBy(F.desc(\"reason_end\")).show()\n", " # skip_reason_df = transformed_df.groupBy(\"skipped\", \"reason_end\").count().orderBy(F.desc(\"reason_end\")).show()\n",
" # reason_df = transformed_df.groupBy(\"reason_end\").orderBy(F.desc(\"count\")).show(5)\n", " # reason_df = transformed_df.groupBy(\"reason_end\").orderBy(F.desc(\"count\")).show(5)\n",
" # skip_df = transformed_df.groupBy(\"skipped\").count().orderBy(F.desc(\"count\")).show(5)\n", " # skip_df = transformed_df.groupBy(\"skipped\").count().orderBy(F.desc(\"count\")).show(5)\n",
" skip_join_bool = transformed_df.select(\"reason_end\", \"was_skipped\").distinct().orderBy(\"reason_end\").show(truncate=False))\n", " # skip_join_bool = transformed_df.select(\"reason_end\", \"skipped\").distinct().orderBy(\"reason_end\").show(100)\n",
" skip_reason_bool = (transformed_df.groupBy(\"reason_end\")\n",
" .agg(F.collect_set(\"skipped\").alias(\"skipped_values\"))\n",
" .orderBy(\"reason_end\")\n",
" .show(truncate=False))\n",
" #join(transformed_df.F.col(\"reason_end\"))\n", " #join(transformed_df.F.col(\"reason_end\"))\n",
"\n", "\n",
" # missing_boolean = set(rename_map) - set(df.columns)\n", " # missing_boolean = set(rename_map) - set(df.columns)\n",

42
spotify_silver_gold.ipynb
View File

@@ -72,7 +72,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": 19,
"id": "f7673c72", "id": "f7673c72",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
@@ -87,16 +87,31 @@
"name": "stdout", "name": "stdout",
"output_type": "stream", "output_type": "stream",
"text": [ "text": [
"+----------+--------------+--------------------+---------+-------+-----------------+--------------------+----------+------------+-------+-------+--------------------+-------------------+--------------------+--------------------+--------+----------+------------+------------+\n", "+--------------------+--------------------+-----+\n",
"|album_name| artist_name| track_name|ms_played|offline|offline_timestamp| platform|reason_end|reason_start|shuffle|skipped| track_uri| ts| username| offline_timestamped| time| date|offline_time|offline_date|\n", "| track_name| username|count|\n",
"+----------+--------------+--------------------+---------+-------+-----------------+--------------------+----------+------------+-------+-------+--------------------+-------------------+--------------------+--------------------+--------+----------+------------+------------+\n", "+--------------------+--------------------+-----+\n",
"|Elton John| Elton John| Your Song| 0.42| false| 1665740300104|iOS 16.0 (iPhone1...| endplay| clickrow| false| NULL|spotify:track:38z...|2022-10-14 10:38:46|fm30ajetkgrrkbvae...|+54755-03-18 14:5...|10:38:46|2022-10-14| 14:55:04|+54755-03-18|\n", "| Mujeriego| csm3-us| 93|\n",
"| Lights 10|Ellie Goulding| Your Song| 3.11| false| 1665740325724|iOS 16.0 (iPhone1...| endplay| clickrow| true| NULL|spotify:track:6XK...|2022-10-14 10:42:32|fm30ajetkgrrkbvae...|+54755-03-18 22:0...|10:42:32|2022-10-14| 22:02:04|+54755-03-18|\n", "| Angelito| csm3-us| 16|\n",
"| 19| Adele|Make You Feel My ...| 3.53| false| 1665740551741|iOS 16.0 (iPhone1...| trackdone| clickrow| false| NULL|spotify:track:273...|2022-10-14 10:46:27|fm30ajetkgrrkbvae...|+54755-03-21 12:4...|10:46:27|2022-10-14| 12:49:01|+54755-03-21|\n", "|Navidad en Mi Barrio| csm3-us| 1|\n",
"| May It Be| Enya| May It Be| 0.07| false| 1665740786443|iOS 16.0 (iPhone1...| backbtn| trackdone| true| NULL|spotify:track:1wI...|2022-10-14 10:46:32|fm30ajetkgrrkbvae...|+54755-03-24 06:0...|10:46:32|2022-10-14| 06:00:43|+54755-03-24|\n", "| Como Perro Y Gato| csm3-us| 1|\n",
"| 19| Adele|Make You Feel My ...| 0.13| false| 1665740791987|iOS 16.0 (iPhone1...| fwdbtn| backbtn| true| NULL|spotify:track:273...|2022-10-14 10:46:40|fm30ajetkgrrkbvae...|+54755-03-24 07:3...|10:46:40|2022-10-14| 07:33:07|+54755-03-24|\n", "| The Taste of Ink|fm30ajetkgrrkbvae...| 2|\n",
"+----------+--------------+--------------------+---------+-------+-----------------+--------------------+----------+------------+-------+-------+--------------------+-------------------+--------------------+--------------------+--------+----------+------------+------------+\n", "| Hasta el Amanecer| csm3-us| 4|\n",
"only showing top 5 rows\n" "| Tu Foto| csm3-us| 1|\n",
"|Love Is Not Over ...| csm3-us| 3|\n",
"|Like I'm Gonna Lo...|fm30ajetkgrrkbvae...| 2|\n",
"| Lo Siento BB:/| csm3-us| 5|\n",
"| Superbeast| csm3-us| 1|\n",
"| Amor de Hospital| csm3-us| 11|\n",
"| All Right Now|fm30ajetkgrrkbvae...| 2|\n",
"|Dancing On The Ce...|fm30ajetkgrrkbvae...| 4|\n",
"|Prologue / The Da...|fm30ajetkgrrkbvae...| 4|\n",
"|That Way (feat. J...| csm3-us| 11|\n",
"| Blame| csm3-us| 29|\n",
"| Speak Now|fm30ajetkgrrkbvae...| 2|\n",
"|A Little Respect ...|fm30ajetkgrrkbvae...| 3|\n",
"| Rockstar|fm30ajetkgrrkbvae...| 4|\n",
"+--------------------+--------------------+-----+\n",
"only showing top 20 rows\n"
] ]
}, },
{ {
@@ -115,9 +130,10 @@
"\n", "\n",
" table_name = \"spotify_data\"\n", " table_name = \"spotify_data\"\n",
" silver_df = read_jdbc(spark, table_name)\n", " silver_df = read_jdbc(spark, table_name)\n",
" silver_df.show(1)\n",
"\n", "\n",
" spark.stop()\n", " silver_df.groupBy(\"track_name\", \"username\").count().show()\n",
"\n",
" #spark.stop()\n",
"if __name__ == \"__main__\":\n", "if __name__ == \"__main__\":\n",
" main()" " main()"
] ]