Add black pre-commit hook

ibis-project · Apr 14, 2023 · 85c84ad · 85c84ad
1 parent 87fb358
commit 85c84ad
Show file tree

Hide file tree

Showing 12 changed files with 190 additions and 181 deletions.
diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
@@ -3,3 +3,7 @@ repos:
     rev: 0.6.0
     hooks:
       - id: nbstripout
+  - repo: https://github.com/psf/black
+    rev: 23.3.0
+    hooks:
+      - id: black-jupyter
diff --git a/examples/Reading-Parquet-Files-using-DuckDB.ipynb b/examples/Reading-Parquet-Files-using-DuckDB.ipynb
@@ -41,7 +41,9 @@
    "source": [
     "import ibis\n",
     "\n",
-    "t = ibis.read_parquet(f\"s3://gbif-open-data-us-east-1/occurrence/2023-04-01/occurrence.parquet/000000\")\n",
+    "t = ibis.read_parquet(\n",
+    "    f\"s3://gbif-open-data-us-east-1/occurrence/2023-04-01/occurrence.parquet/000000\"\n",
+    ")\n",
     "t"
    ]
   },
@@ -66,11 +68,23 @@
    },
    "outputs": [],
    "source": [
-    "cols = ['gbifid', 'datasetkey', 'occurrenceid', 'kingdom',\n",
-    "        'phylum', 'class', 'order', 'family', 'genus',\n",
-    "        'species', 'day', 'month', 'year']\n",
+    "cols = [\n",
+    "    \"gbifid\",\n",
+    "    \"datasetkey\",\n",
+    "    \"occurrenceid\",\n",
+    "    \"kingdom\",\n",
+    "    \"phylum\",\n",
+    "    \"class\",\n",
+    "    \"order\",\n",
+    "    \"family\",\n",
+    "    \"genus\",\n",
+    "    \"species\",\n",
+    "    \"day\",\n",
+    "    \"month\",\n",
+    "    \"year\",\n",
+    "]\n",
     "\n",
-    "t.select(cols).filter(t['family'].isin(['Corvidae'])).limit(5).execute()"
+    "t.select(cols).filter(t[\"family\"].isin([\"Corvidae\"])).limit(5).execute()"
    ]
   },
   {
@@ -111,7 +125,9 @@
    },
    "outputs": [],
    "source": [
-    "t = ibis.read_parquet(f\"s3://gbif-open-data-us-east-1/occurrence/2023-04-01/occurrence.parquet/*\")"
+    "t = ibis.read_parquet(\n",
+    "    f\"s3://gbif-open-data-us-east-1/occurrence/2023-04-01/occurrence.parquet/*\"\n",
+    ")"
    ]
   },
   {
@@ -132,11 +148,11 @@
    "outputs": [],
    "source": [
     "df = (\n",
-    "    t.select(['gbifid', 'family', 'species'])\n",
-    "    .filter(t['family'].isin(['Corvidae']))\n",
+    "    t.select([\"gbifid\", \"family\", \"species\"])\n",
+    "    .filter(t[\"family\"].isin([\"Corvidae\"]))\n",
     "    # Here we limit by 10,000 to fetch a quick batch of results\n",
     "    .limit(10000)\n",
-    "    .group_by('species')\n",
+    "    .group_by(\"species\")\n",
     "    .count()\n",
     "    .execute()\n",
     ")\n",

diff --git a/examples/Substrait.ipynb b/examples/Substrait.ipynb
@@ -49,18 +49,18 @@
     "class DatabaseServer:\n",
     "    DB_NAME = \"palmer_penguins.ddb\"\n",
     "    DB_URL = \"https://storage.googleapis.com/ibis-tutorial-data/palmer_penguins.ddb\"\n",
-    "    \n",
+    "\n",
     "    def __init__(self):\n",
     "        if not os.path.exists(self.DB_NAME):\n",
     "            urlretrieve(self.DB_URL, self.DB_NAME)\n",
     "        self.db = duckdb.connect(self.DB_NAME)\n",
     "        self.db.install_extension(\"substrait\")\n",
     "        self.db.load_extension(\"substrait\")\n",
-    "    \n",
+    "\n",
     "    def execute(self, substrait):\n",
     "        result = self.db.from_substrait(substrait)\n",
     "        return result.fetchall()\n",
-    "        \n",
+    "\n",
     "\n",
     "db_server = DatabaseServer()"
    ]
@@ -84,7 +84,7 @@
     "from ibis.expr.datatypes.core import Float64, Int64, String\n",
     "\n",
     "table = ibis.table(\n",
-    "    name=\"penguins\", \n",
+    "    name=\"penguins\",\n",
     "    schema=[\n",
     "        (\"species\", String()),\n",
     "        (\"island\", String()),\n",
@@ -93,8 +93,8 @@
     "        (\"flipper_length_mm\", Int64()),\n",
     "        (\"body_mass_g\", Int64()),\n",
     "        (\"sex\", String()),\n",
-    "        (\"year\", Int64)\n",
-    "    ]\n",
+    "        (\"year\", Int64),\n",
+    "    ],\n",
     ")\n",
     "\n",
     "print(table)"
@@ -122,12 +122,7 @@
     "\n",
     "compiler = SubstraitCompiler()\n",
     "\n",
-    "query = (\n",
-    "    table\n",
-    "    .select(_.species)\n",
-    "    .group_by(_.species)\n",
-    "    .agg(count=_.species.count())\n",
-    ")\n",
+    "query = table.select(_.species).group_by(_.species).agg(count=_.species.count())\n",
     "\n",
     "substrait_plan = compiler.compile(query)\n",
     "\n",
@@ -172,8 +167,7 @@
    "outputs": [],
    "source": [
     "query = (\n",
-    "    table\n",
-    "    .select(_.island, _.species)\n",
+    "    table.select(_.island, _.species)\n",
     "    .group_by([_.island, _.species])\n",
     "    .agg(num=_.species.count())\n",
     "    .order_by([ibis.asc(_.island), ibis.asc(_.species)])\n",
@@ -198,8 +192,7 @@
    "outputs": [],
    "source": [
     "query = (\n",
-    "    table\n",
-    "    .select(_.island, _.species, _.body_mass_g)\n",
+    "    table.select(_.island, _.species, _.body_mass_g)\n",
     "    .group_by([_.island, _.species])\n",
     "    .agg(num=_.species.count(), avg_weight=_.body_mass_g.mean())\n",
     "    .order_by([ibis.asc(_.island), ibis.asc(_.species)])\n",

diff --git a/examples/clickhouse-hackernews.ipynb b/examples/clickhouse-hackernews.ipynb
@@ -25,7 +25,7 @@
     "import ibis\n",
     "from ibis import _\n",
     "\n",
-    "ibis.options.interactive = True "
+    "ibis.options.interactive = True"
    ]
   },
   {
@@ -46,10 +46,7 @@
    "outputs": [],
    "source": [
     "con = ibis.clickhouse.connect(\n",
-    "    host=\"play.clickhouse.com\", \n",
-    "    port=9440, \n",
-    "    user=\"play\", \n",
-    "    secure=True\n",
+    "    host=\"play.clickhouse.com\", port=9440, user=\"play\", secure=True\n",
     ")"
    ]
   },
@@ -131,9 +128,9 @@
    "source": [
     "top_posts_by_score = (\n",
     "    t.filter(_.title != \"\")\n",
-    "     .select(\"title\", \"score\")\n",
-    "     .order_by(ibis.desc(\"score\"))\n",
-    "     .limit(5)\n",
+    "    .select(\"title\", \"score\")\n",
+    "    .order_by(ibis.desc(\"score\"))\n",
+    "    .limit(5)\n",
     ")\n",
     "\n",
     "top_posts_by_score"
@@ -163,11 +160,7 @@
    "outputs": [],
    "source": [
     "top_commenters = (\n",
-    "    t.filter(_.by != \"\")\n",
-    "     .group_by(\"by\")\n",
-    "     .count()\n",
-    "     .order_by(ibis.desc(\"count\"))\n",
-    "     .limit(5)\n",
+    "    t.filter(_.by != \"\").group_by(\"by\").count().order_by(ibis.desc(\"count\")).limit(5)\n",
     ")\n",
     "\n",
     "top_commenters"
@@ -214,10 +207,10 @@
    "source": [
     "top_commenters_by_score = (\n",
     "    t.filter(_.by != \"\")\n",
-    "     .group_by(\"by\")\n",
-    "     .agg(total_score=_.score.sum())\n",
-    "     .order_by(ibis.desc(\"total_score\"))\n",
-    "     .limit(5)\n",
+    "    .group_by(\"by\")\n",
+    "    .agg(total_score=_.score.sum())\n",
+    "    .order_by(ibis.desc(\"total_score\"))\n",
+    "    .limit(5)\n",
     ")\n",
     "\n",
     "top_commenters_by_score"

diff --git a/tutorial/01-Introduction-to-Ibis.ipynb b/tutorial/01-Introduction-to-Ibis.ipynb
@@ -37,6 +37,7 @@
    "outputs": [],
    "source": [
     "from tutorial_utils import setup\n",
+    "\n",
     "setup()"
    ]
   },
@@ -108,7 +109,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "connection = ibis.sqlite.connect('geography.db')"
+    "connection = ibis.sqlite.connect(\"geography.db\")"
    ]
   },
   {
@@ -150,7 +151,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "countries = connection.table('countries')"
+    "countries = connection.table(\"countries\")"
    ]
   },
   {
@@ -184,7 +185,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "countries['name', 'continent', 'population']"
+    "countries[\"name\", \"continent\", \"population\"]"
    ]
   },
   {
@@ -219,7 +220,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "countries['name', 'continent', 'population'].limit(3)"
+    "countries[\"name\", \"continent\", \"population\"].limit(3)"
    ]
   },
   {
@@ -239,7 +240,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "countries[['continent']].distinct()"
+    "countries[[\"continent\"]].distinct()"
    ]
   },
   {
@@ -256,7 +257,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "countries['continent'] == 'AS'"
+    "countries[\"continent\"] == \"AS\""
    ]
   },
   {
@@ -274,8 +275,8 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "asian_countries = countries['name', 'continent', 'population'].filter(\n",
-    "    countries['continent'] == 'AS'\n",
+    "asian_countries = countries[\"name\", \"continent\", \"population\"].filter(\n",
+    "    countries[\"continent\"] == \"AS\"\n",
     ")\n",
     "asian_countries"
    ]
@@ -310,7 +311,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "asian_countries.order_by('population').limit(10)"
+    "asian_countries.order_by(\"population\").limit(10)"
    ]
   },
   {
@@ -329,7 +330,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "asian_countries.order_by(ibis.desc('population')).limit(10)"
+    "asian_countries.order_by(ibis.desc(\"population\")).limit(10)"
    ]
   },
   {