Skip to content

Commit

Permalink
ML/AutoML: Harmonize notebooks, and easier connection string handling
Browse files Browse the repository at this point in the history
  • Loading branch information
@ckurze @amotl
ckurze authored and amotl committed Dec 2, 2023
1 parent 637960d commit d64a84e
Show file tree
Hide file tree
Showing 2 changed files with 69 additions and 28 deletions.
87 changes: 62 additions & 25 deletions topic/machine-learning/automl/automl_classification_with_pycaret.ipynb
View file
Original file line number Diff line number Diff line change
Expand Up @@ -115,12 +115,27 @@
"source": [
"## Getting started\n",
"\n",
"First, install the required dependencies. \n",
"\n",
"```bash\n",
"pip install -r requirements.txt\n",
"```\n",
"First, install the required dependencies. "
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"#!pip install -r requirements.txt\n",
"\n",
"# In an environment like Google Colab, please use the abolute path of requirements.txt\n",
"# Note: Some inconsistencies of dependencies might get reported. They can usually be ignored.\n",
"# Restart the runtime, if asked by Colab.\n",
"#!pip install -r https://raw.githubusercontent.com/crate/cratedb-examples/main/topic/machine-learning/automl/requirements.txt"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**Note:** As of time of this writing, PyCaret requires Python 3.8, 3.9 or 3.10.\n",
"\n",
"Second, you will need a CrateDB instance to store and serve the data. The easiest\n",
Expand All @@ -131,31 +146,53 @@
"create an `.env` file with the following content:\n",
"\n",
"```env\n",
"CRATE_HOST=<your-crate-host> # set this to localhost if you're running crate locally\n",
"CRATE_USER=<your-crate-user> # set this to crate if you're running crate locally\n",
"CRATE_PASSWORD=<your-crate-password> # set this to \"\" if you're running crate locally\n",
"CRATE_SSL=true # set this to false if you're running crate locally\n",
"# use this string for a connection to CrateDB Cloud\n",
"CONNECTION_STRING=crate://username:password@hostname/?ssl=true \n",
"\n",
"# use this string for a local connection to CrateDB\n",
"# CONNECTION_STRING=crate://crate@localhost/?ssl=false\n",
"```\n",
"\n",
"You can find your CrateDB credentials in the [CrateDB Cloud Console].\n",
"\n",
"[CrateDB Cloud Console]: https://cratedb.com/docs/cloud/en/latest/reference/overview.html#cluster\n",
"[deploy a cluster]: https://cratedb.com/docs/cloud/en/latest/tutorials/deploy/stripe.html#deploy-cluster\n",
"\n",
"### Creating demo data\n",
"[deploy a cluster]: https://cratedb.com/docs/cloud/en/latest/tutorials/deploy/stripe.html#deploy-cluster"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"\n",
"For convenience, this notebook comes with an accompanying CSV dataset which you\n",
"can quickly import into the database. Upload the CSV file to your CrateDB cloud\n",
"cluster, as described [here](https://cratedb.com/docs/cloud/en/latest/reference/overview.html#import).\n",
"To follow this notebook, choose `pycaret_churn` for your table name.\n",
"# For CrateDB Cloud, use:\n",
"CONNECTION_STRING = os.environ.get(\n",
" \"CRATEDB_CONNECTION_STRING\",\n",
" \"crate://username:password@hostname/?ssl=true\",\n",
")\n",
"\n",
"This will automatically create a new database table and import the data."
"# For an self-deployed CrateDB, e.g. via Docker, please use:\n",
"# CONNECTION_STRING = os.environ.get(\n",
"# \"CRATEDB_CONNECTION_STRING\",\n",
"# \"crate://crate@localhost/?ssl=false\",\n",
"# )"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Creating demo data\n",
"\n",
"For convenience, this notebook comes with an accompanying CSV dataset which you\n",
"can quickly import into the database. Upload the CSV file to your CrateDB cloud\n",
"cluster, as described [here](https://cratedb.com/docs/cloud/en/latest/reference/overview.html#import).\n",
"To follow this notebook, choose `pycaret_churn` for your table name.\n",
"\n",
"This will automatically create a new database table and import the data.\n",
"\n",
"### Alternative data import using code\n",
"\n",
"If you prefer to use code to import your data, please execute the following lines which read the CSV\n",
Expand All @@ -175,8 +212,7 @@
"if os.path.exists(\".env\"):\n",
" dotenv.load_dotenv(\".env\", override=True)\n",
"\n",
"dburi = f\"crate://{os.environ['CRATE_USER']}:{os.environ['CRATE_PASSWORD']}@{os.environ['CRATE_HOST']}:4200?ssl={os.environ['CRATE_SSL']}\"\n",
"engine = sa.create_engine(dburi, echo=os.environ.get('DEBUG'))\n",
"engine = sa.create_engine(CONNECTION_STRING, echo=os.environ.get('DEBUG'))\n",
"df = pd.read_csv(\"https://github.com/crate/cratedb-datasets/raw/main/machine-learning/automl/churn-dataset.csv\")\n",
"\n",
"with engine.connect() as conn:\n",
Expand Down Expand Up @@ -214,8 +250,7 @@
"if os.path.exists(\".env\"):\n",
" dotenv.load_dotenv(\".env\", override=True)\n",
"\n",
"dburi = f\"crate://{os.environ['CRATE_USER']}:{os.environ['CRATE_PASSWORD']}@{os.environ['CRATE_HOST']}:4200?ssl={os.environ['CRATE_SSL']}\"\n",
"engine = sa.create_engine(dburi, echo=os.environ.get('DEBUG'))\n",
"engine = sa.create_engine(CONNECTION_STRING, echo=os.environ.get('DEBUG'))\n",
"\n",
"with engine.connect() as conn:\n",
" with conn.execute(sa.text(\"SELECT * FROM pycaret_churn\")) as cursor:\n",
Expand All @@ -224,7 +259,7 @@
"# We set the MLFLOW_TRACKING_URI to our CrateDB instance. We'll see later why\n",
"os.environ[\n",
" \"MLFLOW_TRACKING_URI\"\n",
"] = f\"{dburi}&schema=mlflow\""
"] = f\"{CONNECTION_STRING}&schema=mlflow\""
]
},
{
Expand Down Expand Up @@ -966,8 +1001,10 @@
"# - \"n_select\" defines how many models are selected.\n",
"# - \"exclude\" defines which models are excluded from the comparison.\n",
"\n",
"# Note: This is only relevant if we are executing automated tests\n",
"if \"PYTEST_CURRENT_TEST\" in os.environ:\n",
" best_models = compare_models(sort=\"AUC\", include=[\"lr\", \"knn\"], n_select=3)\n",
"# If we are not in an automated test, compare the available models\n",
"else:\n",
" # For production scenarios, it might be worth to include \"lightgbm\" again.\n",
" best_models = compare_models(sort=\"AUC\", exclude=[\"lightgbm\"], n_select=3)"
Expand Down Expand Up @@ -3406,7 +3443,7 @@
"source": [
"os.environ[\n",
" \"MLFLOW_TRACKING_URI\"\n",
"] = f\"crate://{os.environ['CRATE_USER']}:{os.environ['CRATE_PASSWORD']}@{os.environ['CRATE_HOST']}:4200?ssl={os.environ['CRATE_SSL']}&schema=mlflow\""
"] = f\"{CONNECTION_STRING}&schema=mlflow\""
]
},
{
Expand Down Expand Up @@ -3484,7 +3521,7 @@
],
"metadata": {
"kernelspec": {
"display_name": "crate",
"display_name": "Python 3 (ipykernel)",
"language": "python",
"name": "python3"
},
Expand All @@ -3498,7 +3535,7 @@
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.10.0"
"version": "3.11.4"
}
},
"nbformat": 4,
Expand Down
10 changes: 7 additions & 3 deletions topic/machine-learning/automl/automl_timeseries_forecasting_with_pycaret.ipynb
View file
Original file line number Diff line number Diff line change
Expand Up @@ -112,13 +112,15 @@
},
{
"cell_type": "code",
"execution_count": null,
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"#!pip install -r requirements.txt\n",
"\n",
"# In an environment like Google Colab, please use the abolute path of requirements.txt:\n",
"# In an environment like Google Colab, please use the abolute path of requirements.txt\n",
"# Note: Some inconsistencies of dependencies might get reported. They can usually be ignored.\n",
"# Restart the runtime, if asked by Colab.\n",
"#!pip install -r https://raw.githubusercontent.com/crate/cratedb-examples/main/topic/machine-learning/automl/requirements.txt"
]
},
Expand Down Expand Up @@ -1080,10 +1082,12 @@
"# all available models are included by default)\n",
"# - \"fold\" defines the number of folds to use for cross-validation.\n",
"\n",
"# Note: This is only relevant if we are executing automated tests\n",
"if \"PYTEST_CURRENT_TEST\" in os.environ:\n",
" best_models = compare_models(sort=\"MASE\",\n",
" include=[\"ets\", \"et_cds_dt\", \"naive\"],\n",
" n_select=3)\n",
"# If we are not in an automated test, compare all available models\n",
"else:\n",
" best_models = compare_models(sort=\"MASE\", n_select=3)"
]
Expand Down Expand Up @@ -1831,7 +1835,7 @@
"The missing step is to identify the best model from all the conducted experiments.\n",
"This is done by simply looking the the Mean MASE outputs of all the model training\n",
"and tuning steps above. Looking at all the outputs, the best performing model is\n",
"the blended model with a MASE of 0.7783."
"the blended model with a MASE of approx. 0.77."
]
},
{
Expand Down

0 comments on commit d64a84e

Please sign in to comment.