.. DO NOT EDIT.
.. THIS FILE WAS AUTOMATICALLY GENERATED BY SPHINX-GALLERY.
.. TO MAKE CHANGES, EDIT THE SOURCE PYTHON FILE:
.. "examples_gallery/plot_arps_fit.py"
.. LINE NUMBERS ARE GIVEN BELOW.

.. only:: html

    .. note::
        :class: sphx-glr-download-link-note

        :ref:`Go to the end <sphx_glr_download_examples_gallery_plot_arps_fit.py>`
        to download the full example code.

.. rst-class:: sphx-glr-example-title

.. _sphx_glr_examples_gallery_plot_arps_fit.py:


=================
Fit an Arps curve
=================

Here we fit a curve to data using the low-level API and plot it.

.. GENERATED FROM PYTHON SOURCE LINES 9-80



.. image-sg:: /examples_gallery/images/sphx_glr_plot_arps_fit_001.png
   :alt: plot arps fit
   :srcset: /examples_gallery/images/sphx_glr_plot_arps_fit_001.png
   :class: sphx-glr-single-img


.. rst-class:: sphx-glr-script-out

 .. code-block:: none

    Optimal parameters are: [ 6.658  2.188 -5.326]






|

.. code-block:: Python


    import matplotlib.pyplot as plt
    import numpy as np
    from scipy.optimize import minimize

    from dca.datasets import load_monthly_sodir_production
    from dca.decline_curve_analysis import Arps, CurveLoss

    # Load data to fit to
    df = (
        load_monthly_sodir_production()
        .loc[lambda df: df["prfInformationCarrier"] == "STATFJORD", :]
        .groupby("prfYear")["prfPrdOilNetMillSm3"]
        .sum()
    )


    # We fit the log-data to a log-model, and we shift the x-values
    x = df.index - df.index.min() + 0.5
    y = np.log(df.values)


    def log_arps(t, theta1, theta2, theta3):
        """Curve function to fit to."""
        return Arps(theta1, theta2, theta3).eval_log(t)


    def log_arps_grad(t, theta1, theta2, theta3):
        """Evaluate the gradient of the log of the Arps model."""
        return Arps(theta1, theta2, theta3).eval_grad_log(t)


    # Prior mean and optimization first guess for parameters
    mu = np.array([10, 2, 0])
    # Controls prior strength
    alpha = 1e-2
    precision_matrix = np.eye(3) * alpha
    loss_function = CurveLoss(
        mu=mu,
        precision_matrix=precision_matrix,
        curve_func=log_arps,
        curve_func_grad=log_arps_grad,
        half_life=24,
        p=1.5,
    )

    # Optimize the function
    optimization_result = minimize(
        loss_function,
        x0=mu,
        method="BFGS",
        jac=loss_function.grad,  # Use gradient information
        args=(
            x,
            y,
        ),
    )
    assert optimization_result.success
    optimal_parameters = optimization_result.x
    print(f"Optimal parameters are: {optimal_parameters.round(3)}")


    # Plot the results
    plt.plot(x, y, "-o", label="Data")
    curve = Arps(*optimal_parameters)
    plt.plot(x, curve.eval_log(x), label="Arps curve fit")
    plt.xlabel("Months since production start")
    plt.ylabel("log Production")
    plt.grid(True, ls="--", alpha=0.33)
    plt.legend()
    plt.show()


.. rst-class:: sphx-glr-timing

   **Total running time of the script:** (0 minutes 0.088 seconds)


.. _sphx_glr_download_examples_gallery_plot_arps_fit.py:

.. only:: html

  .. container:: sphx-glr-footer sphx-glr-footer-example

    .. container:: sphx-glr-download sphx-glr-download-jupyter

      :download:`Download Jupyter notebook: plot_arps_fit.ipynb <plot_arps_fit.ipynb>`

    .. container:: sphx-glr-download sphx-glr-download-python

      :download:`Download Python source code: plot_arps_fit.py <plot_arps_fit.py>`

    .. container:: sphx-glr-download sphx-glr-download-zip

      :download:`Download zipped: plot_arps_fit.zip <plot_arps_fit.zip>`


.. only:: html

 .. rst-class:: sphx-glr-signature

    `Gallery generated by Sphinx-Gallery <https://sphinx-gallery.github.io>`_