.. _examples.business.spotify:

Predicting Popularity on Spotify
=================================

This example uses the publicly-available Spotify from Kaggle to predict the popularity of Polish songs and artists on Spotify. We'll also use a model to group artists together based on how similar their songs are.

.. note:: We are only using polish artists and a subset of the tracks dataset filtered by a handful of artists.

The ``tracks`` dataset (``tracks.csv``) have the following features:

Id represents the Id of the track generated by Spotify

Numerical:

- **acousticness** (range: ``[0,1]``)
- **danceability** (range: ``[0,1]``)
- **energy** (range: ``[0,1]``)
- **duration_ms** (range: ``[200000,300000]``)
- **instrumentalness** (range: ``[0,1]``)
- **valence** (range: ``[0,1]``)
- **popularity** (range: ``[0,100]``)
- **tempo** (range: ``[50,150]``)
- **liveness** (range: ``[0,1]``)
- **loudness** (range: ``[-60,0]``)
- **speechiness** (range: ``[0,1]``)

Dummy:

- **mode**: (``0 = Minor``, ``1 = Major``)
- **explicit**: (0 = No explicit content and 1 = Explicit content)

Categorical:

- **key**: keys on an octave encoded as integers in range ``[0,11]`` (``C = 0``, ``C = 1``, etc.)
- **timesignature**: predicted time signature.
- **artists**: list of contributing artists.
- **artists**: list of IDs of contributing artists.
- **release_date**: date of release (yyyy-mm-dd).
- **name**: track name.

The ``artists`` dataset (``artists.csv``) has the following features:

- **id**: ID of the artist.
- **name**: artist name.
- **followers**: how many followers the artist has.
- **popularity**: popularity of the artists based on their tracks.
- **genres**: list of genres covered by the artist's tracks.

We will follow the data science cycle (Data Exploration - Data Preparation - Data Modeling - Model Evaluation - Model Deployment) to solve this problem.

Initialization
---------------

This example uses the following version of vastorbit:

.. ipython:: python
    
    import vastorbit as vo

    vo.__version__

Start by importing vastorbit and loading the SQL extension, which allows you to query the VAST DataBase with SQL.

.. ipython:: python

    %load_ext vastorbit.sql


Connect to VAST. This example uses an existing connection called ``VASTDSN``. 
For details on how to create a connection, see the :ref:`connection` tutorial.
You can skip the below cell if you already have an established connection.

.. code-block:: python
    
    vo.connect("VASTDSN")

Data Loading
-------------

Load the datasets into the :py:mod:`~vastorbit.VastFrame` with :py:func:`~vastorbit.read_csv` and then view them with :py:func:`~vastorbit.VastFrame.head`.

.. code-block::

    # load datasets as VastFrame objects
    artists = vo.read_csv("artists.csv")
    tracks  = vo.read_csv("tracks.csv")

    # Display
    artists

.. ipython:: python
    :suppress:

    try:
        artists = vo.read_csv(
            "/Users/badr.ouali/Documents/VastOrbit-master/docs/source/_static/website/examples/data/spotify/artists.csv",
        )
    except:
        artists = vo.VastFrame("artists")
    res = artists
    html_file = open("/Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_artists_table.html", "w")
    html_file.write(res._repr_html_())
    html_file.close()

.. raw:: html
    :file: /Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_artists_table.html

.. code-block::

    tracks

.. ipython:: python
    :suppress:
    
    try:
        tracks = vo.read_csv("/Users/badr.ouali/Documents/VastOrbit-master/docs/source/_static/website/examples/data/spotify/tracks.csv")
    except:
        tracks = vo.VastFrame("tracks")
    res = tracks
    html_file = open("/Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_tracks_table.html", "w")
    html_file.write(res._repr_html_())
    html_file.close()

.. raw:: html
    :file: /Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_tracks_table.html

.. warning::
    
    This example uses a sample dataset. It only includes Polish artists. For the full analysis, you should consider using the complete dataset.

Data Exploration 
-----------------

We can visualize the top ``60`` most-followed Polish artists with a bar chart.

.. code-block:: python

    # make a chart of the top 50 most-followed Polish artists
    artists.bar(
        ["name"], 
        method = "mean",
        of = "followers",
        max_cardinality = 50,
        width = 800,
    )

.. ipython:: python
    :suppress:

    vo.set_option("plotting_lib","plotly")
    fig = artists.bar(
        ["name"], 
        method = "mean",
        of = "followers",
        max_cardinality = 50,
        width = 800,
    )
    fig.write_html("/Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_polish_followers_bar.html")

.. raw:: html
    :file: /Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_polish_followers_bar.html

We can do the same with the most popular tracks. For example, we can graph Monika Brodka's most popular tracks like so:

.. code-block::

    # find Monika Brodka's songs
    brodka_tracks = tracks.search("LOWER(artists) like '%brodka%'")

    # plot Brodka's tracks ordered by popularity
    brodka_tracks.bar(
        ["name"], 
        method = "mean",
        of = "popularity",
        max_cardinality = 25,
        width = 800,
    )

.. ipython:: python
    :suppress:
    :okwarning:

    brodka_tracks = tracks.search("LOWER(artists) like '%brodka%'")
    fig = brodka_tracks.bar(
        ["name"], 
        method = "mean",
        of = "popularity",
        max_cardinality = 25,
        width = 800,
    )
    fig.write_html("/Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_brodka_popularity_bar.html")

.. raw:: html
    :file: /Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_brodka_popularity_bar.html

To get an idea of what makes Monika Brodka's songs popular, let's create a boxplot of the numerical feature distribution of her tracks.

.. code-block::

    ## list of the relevant numerical features
    numerical_features = [
        "danceability", 
        "energy", 
        "speechiness", 
        "acousticness", 
        "instrumentalness", 
        "valence", 
        "liveness",
    ]

    # create a boxplot of the above features
    brodka_tracks.boxplot(columns = numerical_features)

.. ipython:: python
    :suppress:
    :okwarning:

    ## list of the relevant numerical features
    numerical_features = [
        "danceability", 
        "energy", 
        "speechiness", 
        "acousticness", 
        "instrumentalness", 
        "valence", 
        "liveness",
    ]

    # create a boxplot of the above features
    fig = brodka_tracks.boxplot(columns = numerical_features)
    fig.write_html("/Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_boxplot.html")

.. raw:: html
    :file: /Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_boxplot.html

Timing is a classic factor for success, so let's look at the popularity of Monika's songs over time with a smooth curve.

.. code-block:: 

    # extract year from the date
    brodka_tracks["release_year"] = "YEAR(CAST(release_date AS DATE))"

    # smooth the popularity using rolling mean
    brodka_tracks.rolling(
        func = "mean",
        columns = "popularity",
        window = (-3, 3),
        order_by = "release_year",
        name = "smoothed_popularity",
    )

    # plot the smoothed curve for popularity of her songs
    brodka_tracks.plot(ts = "release_date", columns=["smoothed_popularity"]) 

.. ipython:: python
    :okwarning:
    :suppress:

    # extract year from the date
    brodka_tracks["release_year"] = "YEAR(CAST(release_date AS DATE))"

    # smooth the popularity using rolling mean
    brodka_tracks.rolling(
        func = "mean",
        columns = "popularity",
        window = (-3, 3),
        order_by = "release_year",
        name = "smoothed_popularity",
    )

    # plot the smoothed curve for popularity of her songs
    fig = brodka_tracks.plot(ts = "release_date", columns = ["smoothed_popularity"]) 
    fig.write_html("/Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_brodka_release_plot.html")

.. raw:: html
    :file: /Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_brodka_release_plot.html

Numerical-feature Analysis
---------------------------

Bringing it all together, let's try to get an idea of how these numerical 
features change and correlate with each other in Monika's most popular songs.

.. code-block::

    # extract year from date
    tracks["release_year"] = "YEAR(CAST(release_date AS DATE))"

    # get the average of numerical features during the year
    yearly_aggs = tracks.groupby(
        "release_year", [
            "AVG(danceability) as danceability",
            "AVG(energy) as energy", 
            "AVG(speechiness) AS speechiness", 
            "AVG(acousticness) AS acousticness", 
            "AVG(instrumentalness) AS instrumentalness",
            "AVG(valence) AS valence", 
            "AVG(liveness) AS liveness",
        ]
    )


    # plot the cures for numerical features along the different years
    yearly_aggs.plot(
        ts = "release_year", 
        columns = numerical_features,
    )

.. ipython:: python
    :suppress:
    :okwarning:

    tracks['release_year'] = "YEAR(CAST(release_date AS DATE))"
    yearly_aggs = tracks.groupby(
        "release_year", [
            "AVG(danceability) as danceability",
            "AVG(energy) as energy", 
            "AVG(speechiness) AS speechiness", 
            "AVG(acousticness) AS acousticness", 
            "AVG(instrumentalness) AS instrumentalness",
            "AVG(valence) AS valence", 
            "AVG(liveness) AS liveness",
        ]
    )
    fig = yearly_aggs.plot(
        ts = "release_year", 
        columns = numerical_features,
    )
    fig.write_html("/Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_brodka_release_plot.html")

.. raw:: html
    :file: /Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_brodka_release_plot.html

.. code-block::

    # correlation of numerical features
    tracks[tracks[numerical_features]].corr()

.. ipython:: python
    :suppress:
    :okwarning:

    fig = tracks[tracks[numerical_features]].corr()
    fig.write_html("/Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_tracks_corr.html")

.. raw:: html
    :file: /Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_tracks_corr.html

Feature Engineering
--------------------

To expand our analysis, let's take into account some descriptive features. Since our goal is to predict popularity, some useful features might be:

- number of followers
- popularity for the artist of the track
- the number of artists per track

Additionally, we manipulate our data a bit to make things easier later on:

- converting the duration unit from ``ms`` to ``minute``
- extracting the year from the date.

.. code-block:: python

    %%sql
    DROP TABLE IF EXISTS polish_tracks;
    CREATE TABLE polish_tracks AS
    SELECT * FROM tracks 
    WHERE id_artists IN (SELECT t.id_artists FROM tracks t JOIN artists p
                        ON t.id_artists LIKE '%' || p.id || '%');
    CREATE TABLE polish_tracks_clean AS
    SELECT 
        x.*, 
        x.duration_ms / 60000 AS duration_minute,
        CAST(x.release_date AS date) AS release_year,
        y.followers AS artists_followers,
        y.popularity AS artist_popularity
    FROM polish_tracks AS x LEFT JOIN artists AS y
    ON x.id_artists LIKE '%' || y.id || '%';

.. ipython:: python
    :okwarning:
    :suppress:

    from vastorbit._utils._sql._sys import _executeSQL

    _executeSQL(
        """
        DROP TABLE IF EXISTS polish_tracks
        """
    )
    _executeSQL(
        """
        CREATE TABLE polish_tracks AS
        SELECT * FROM tracks 
        WHERE id_artists IN (SELECT t.id_artists FROM tracks t JOIN artists p
                            ON t.id_artists LIKE '%' || p.id || '%')
        """
    )
    _executeSQL(
        """
        CREATE TABLE polish_tracks_clean AS
        SELECT 
            x.*, 
            x.duration_ms / 60000 AS duration_minute,
            CAST(x.release_date AS DATE) AS release_year,
            y.followers AS artists_followers,
            y.popularity AS artist_popularity
        FROM polish_tracks AS x LEFT JOIN artists AS y
        ON x.id_artists LIKE '%' || y.id || '%'
        """
    )

.. code-block:: python

    polish_tracks = vo.VastFrame("polish_tracks_clean")

    # count the number of artists per track
    polish_tracks.regexp(
        column = "artists", 
        pattern = ",", 
        method = "count",
        name = "nb_singers",
    )
    polish_tracks["nb_singers"].add(1)

.. ipython:: python
    :suppress:
    :okwarning:

    polish_tracks = vo.VastFrame("polish_tracks_clean")

    # count the number of artists per track
    polish_tracks.regexp(
        column = "artists", 
        pattern = ",", 
        method = "count",
        name = "nb_singers",
    )
    polish_tracks["nb_singers"].add(1)
    res = polish_tracks
    html_file = open("/Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_polish_tracks_clean_table.html", "w")
    html_file.write(res._repr_html_())
    html_file.close()

.. raw:: html
    :file: /Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_polish_tracks_clean_table.html

Define a list of predictors and the response, and then save the normalized version of the final dataset to the database.

.. code-block:: 

    # define predictors and response
    predictors = [
        "duration_minute", 
        # "release_year", 
        "danceability", 
        "energy", 
        "loudness",
        "speechiness", 
        "acousticness", 
        "instrumentalness", 
        "liveness", 
        "valence", 
        "artists_followers", 
        "artist_popularity", 
        "nb_singers",
    ]
    response = "popularity"
    polish_tracks.normalize(
        method = "minmax",
        columns = predictors,
    )
    # save the final dataset to the database
    vo.drop("polish_tracks_data_final")
    polish_tracks.to_db("polish_tracks_data_final", relation_type = "table")

.. ipython:: python
    :suppress:
    :okwarning:

    predictors = [
        "duration_minute", 
        # "release_year", 
        "danceability", 
        "energy", 
        "loudness",
        "speechiness", 
        "acousticness", 
        "instrumentalness", 
        "liveness", 
        "valence", 
        "artists_followers", 
        "artist_popularity", 
        "nb_singers",
    ]
    response = "popularity"
    polish_tracks.normalize(
        method = "minmax",
        columns = predictors,
    )
    vo.drop("polish_tracks_data_final")
    polish_tracks.to_db("polish_tracks_data_final", relation_type = "table")

Machine Learning
-----------------

We can use :py:mod:`~vastorbit.machine_learning.vast.automl.AutoML` to easily get a well-performing model.

.. ipython:: python

    # define a random seed so models tested by AutoML produce consistent results
    vo.set_option("random_state", 2)

:py:mod:`~vastorbit.machine_learning.vast.automl.AutoML` automatically tests several machine learning models and picks the best performing one.

.. ipython:: python
    :okwarning:

    from vastorbit.machine_learning.vast.automl import AutoML

    # define the model
    auto_model = AutoML(
        "automl_spotify_polish",
        estimator = "fast",
        preprocess_data = True,
        stepwise = False,
        cv = 2,
    )

Train the model.

.. ipython:: python
    :okwarning:

    auto_model.fit(
        "polish_tracks_data_final", 
        predictors, 
        response
    )   

.. code-block::

    auto_model.plot()

.. ipython:: python
    :suppress:
    :okwarning:

    fig = auto_model.plot()
    fig.write_html("/Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_automl_plot.html")

.. raw:: html
    :file: /Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_automl_plot.html

Extract the best model according to :py:mod:`~vastorbit.machine_learning.vast.automl.AutoML`. From here, we can look at the model type and its hyperparameters.

.. ipython:: python

    # extract the model type and hyperparameters
    best_model = auto_model.best_model_
    bm_type = best_model._model_type
    hyperparams = best_model.get_params()

    print(bm_type)
    print(hyperparams)

Thanks to :py:mod:`~vastorbit.machine_learning.vast.automl.AutoML`, we know best model type and its hyperparameters. Let's create a new model with this information in mind.

.. code-block:: 

    from vastorbit.machine_learning.vast import LinearRegression

    # define the model
    rf_model = LinearRegression("linear_regression_spotify", **hyperparams)

    # train the model
    rf_model.fit(polish_tracks, predictors, response) 

    # use the model to predict
    rf_model.predict(
        polish_tracks, 
        name = "estimated_popularity",
    )


.. ipython:: python
    :suppress:
    :okwarning:

    from vastorbit.machine_learning.vast import LinearRegression

    # define the model
    rf_model = LinearRegression("linear_regression_spotify", **hyperparams)

    # train the model
    rf_model.fit(polish_tracks, predictors, response) 

    # use the model to predict
    res = rf_model.predict(
        polish_tracks, 
        name = "estimated_popularity",
    )
    html_file = open("/Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_lr_prediction.html", "w")
    html_file.write(res._repr_html_())
    html_file.close()

.. raw:: html
    :file: /Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_lr_prediction.html

View the regression report and the importance of each feature.

.. code-block::

    rf_model.regression_report()

.. ipython:: python
    :suppress:
    :okwarning:

    res = rf_model.regression_report()
    html_file = open("/Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_lr_report.html", "w")
    html_file.write(res._repr_html_())
    html_file.close()

.. raw:: html
    :file: /Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_lr_report.html

.. code-block::

    rf_model.features_importance()

.. ipython:: python
    :suppress:
    :okwarning:

    fig = rf_model.features_importance()
    fig.write_html("/Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_lr_featrures.html")

.. raw:: html
    :file: /Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_lr_featrures.html

To see how our model performs, let's plot the popularity and estimated popularity of songs by other Polish artists like Brodka and Akcent.

.. code-block::

    # results for Brodka
    polish_tracks.search(
            "LOWER(artists) LIKE '%brodka%'",
            usecols = [
                "popularity",
                "name",
                "estimated_popularity",
            ],
        ).plot(
        ts = "name", 
        columns = ["popularity", "estimated_popularity"],
    )

.. ipython:: python
    :suppress:
    :okwarning:

    fig = polish_tracks.search(
            "LOWER(artists) LIKE '%brodka%'",
            usecols = [
                "popularity",
                "name",
                "estimated_popularity",
            ],
        ).plot(
        ts = "name", 
        columns = ["popularity", "estimated_popularity"],
    )
    fig.write_html("/Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_lr_brodaka_predict_plot.html")

.. raw:: html
    :file: /Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_lr_brodaka_predict_plot.html

.. code-block::

    # results for Brodka
    polish_tracks.search(
            "LOWER(artists) LIKE '%akcent%'",
            usecols = [
                "popularity",
                "name",
                "estimated_popularity",
            ],
        ).plot(
        ts = "name",
        columns = [
            "popularity",
            "estimated_popularity",
        ],
    )

.. ipython:: python
    :suppress:
    :okwarning:

    fig = polish_tracks.search(
            "LOWER(artists) LIKE '%akcent%'",
            usecols = [
                "popularity",
                "name",
                "estimated_popularity",
            ],
        ).plot(
        ts = "name",
        columns = [
            "popularity",
            "estimated_popularity",
        ],
    )
    fig.write_html("/Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_lr_akcent_predict_plot.html")

.. raw:: html
    :file: /Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_lr_akcent_predict_plot.html

Group Artists using Track Features
------------------------------------

While our tracks don't have an explicit "genre" feature, we can approximate the effect by grouping artists based on their tracks' numerical features.

Let's start by taking the averages of these numerical features for each artist.

.. code-block::

    # group by artist
    artists_features = polish_tracks.groupby(
        [
            "id_artists",
            "artists",
        ], 
        expr=[
            "AVG(danceability) AS danceability",
            "AVG(energy) AS energy", 
            "AVG(speechiness) AS speechiness", 
            "AVG(acousticness) AS acousticness", 
            "AVG(instrumentalness) AS instrumentalness",
            "AVG(valence) AS valence", 
            "AVG(liveness) AS liveness",
        ]
    )

    # save relation to the database as "artists_features"
    artists_features.to_db("artists_features")

.. ipython:: python
    :suppress:
    :okwarning:

    artists_features = polish_tracks.groupby(
        [
            "id_artists",
            "artists",
        ], 
        expr=[
            "AVG(danceability) AS danceability",
            "AVG(energy) AS energy", 
            "AVG(speechiness) AS speechiness", 
            "AVG(acousticness) AS acousticness", 
            "AVG(instrumentalness) AS instrumentalness",
            "AVG(valence) AS valence", 
            "AVG(liveness) AS liveness",
        ]
    )

    # save relation to the database as "artists_features"
    try:
        artists_features.to_db("artists_features")
    except:
        artists_features = vo.VastFrame("artists_features")
    res = artists_features
    html_file = open("/Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_artists_features.html", "w")
    html_file.write(res._repr_html_())
    html_file.close()

.. raw:: html
    :file: /Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_artists_features.html

Grouping means clustering, so we use an :py:func:`~vastorbit.machine_learning.model_selection.elbow` curve to find a suitable number of clusters.

.. ipython:: python
    :okwarning:

    from vastorbit.machine_learning.model_selection import elbow

    # define numerical features
    preds = [
        "danceability",
        "energy",
        "speechiness",
        "acousticness",
        "instrumentalness",
        "liveness",
        "valence",
    ]

    # elbow curve
    elbow_curve = elbow(
        "artists_features",
        preds,
        n_clusters = (1, 20),
        show = True,
    )

.. code-block:: 

    elbow_curve

.. ipython:: python
    :suppress:
    :okwarning:

    fig = elbow_curve
    fig.write_html("/Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_lr_elbow.html")

.. raw:: html
    :file: /Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_lr_elbow.html

Let's define and use the VAST :py:mod:`~vastorbit.machine_learning.vast.cluster.KMeans` algorithm to create a model that can group artists together.

.. ipython:: python
    :okwarning:

    from vastorbit.machine_learning.vast.cluster import KMeans

    # define k-means
    model = KMeans(
        "KMeans_spotify", 
        n_clusters = 7,
    )

We can train our new model on the ``artists_features`` relation we saved earlier.

.. ipython:: python

    # train the model
    model.fit(
        "artists_features", 
        X = preds,
    )

Plot the result of the k-means algoritm:

.. code-block:: 

    model.plot()

.. ipython:: python
    :suppress:
    :okwarning:

    fig = model.plot()
    fig.write_html("/Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_cluster_plot.html")

.. raw:: html
    :file: /Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_cluster_plot.html

.. ipython:: python

    # predict the genres
    pred_genres = model.predict(
        "artists_features", 
        X = [
            "danceability",
            "energy",
            "speechiness",
            "acousticness",
            "instrumentalness",
            "liveness",
            "valence",
        ], 
        name = "pred_genres",
    )

Let's see how our model groups these artists together:

.. code-block::

    # observe the results
    pred_genres["artists", "pred_genres"].sort({"pred_genres": "desc"})

.. ipython:: python
    :suppress:
    :okwarning:

    res = pred_genres["artists", "pred_genres"].sort({"pred_genres": "desc"})
    html_file = open("/Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_pred_genres.html", "w")
    html_file.write(res._repr_html_())
    html_file.close()

.. raw:: html
    :file: /Users/badr.ouali/Documents/VastOrbit-master/docs/figures/examples_spotify_pred_genres.html

Conclusion
-----------

We were able to predict the popularity Polish songs with a :py:mod:`~vastorbit.machine_learning.vast.ensemble.RandomForestRegressor` model suggested by :py:mod:`~vastorbit.machine_learning.vast.automl.AutoML`. We then created a :py:mod:`~vastorbit.machine_learning.vast.cluster.KMeans` model to group artists into ``genres`` (clusters) based on the feature-commonalities in their tracks.

.. ipython:: python
   :suppress:

   from vastorbit._utils._sql._sys import purge_memory
   purge_memory()