mapreader.process.occlusion_analysis
====================================

.. py:module:: mapreader.process.occlusion_analysis


Classes
-------

.. autoapisummary::

   mapreader.process.occlusion_analysis.OcclusionAnalyzer


Module Contents
---------------

.. py:class:: OcclusionAnalyzer(patch_df, model, transform = 'default', delimiter = ',', device = 'default')

   A class for carrying out occlusion analysis on patches.

   :param patch_df: The DataFrame containing patches and predictions.
   :type patch_df: pd.DataFrame | gpd.GeoDataFrame | str | pathlib.Path
   :param model: The PyTorch model to add to the object.
                 If a string, this should be the path to a model checkpoint.
   :type model: str or nn.Module
   :param transform: The transform to apply to the patches. Options of "default" or a torchvision transform.
                     Default transform
   :type transform: str or callable
   :param delimiter: The delimiter used in the patch_df csv file. By default, ",".
   :type delimiter: str
   :param device: The device to use. By default, "default".
   :type device: str


   .. py:attribute:: loss_fn
      :value: None



   .. py:method:: add_loss_fn(loss_fn = 'cross entropy')

      Add a loss function to the object.

      :param loss_fn: The loss function to use.
                      Can be a string or a torch.nn loss function. Accepted string values are "cross entropy" or "ce" (cross-entropy), "bce" (binary cross-entropy) and "mse" (mean squared error).
                      By default, "cross entropy" is used.
      :type loss_fn: str or torch.nn.modules.loss._Loss



   .. py:method:: run_occlusion(label, sample_size = 10, save = False, path_save = './occlusion_analysis/', block_size = 14)

      Run occlusion analysis on a sample of patches for a given label.

      :param label: The label to run the analysis on.
      :type label: str
      :param sample_size: The number of patches to run the analysis on.
                          By default, 10.
      :type sample_size: int
      :param save: Whether to save the occlusion analysis images. By default, False.
      :type save: bool
      :param path_save: The path to save the occlusion analysis images to. By default, "./occlusion_analysis/".
      :type path_save: str
      :param block_size: The size of the occlusion block. By default, 14.
      :type block_size: int