Visualization

Visualization Module

Functions:

compare_images(title1, im1, title2, im2[, ...])	param im1: image 1.
compare_orig_mask_gt_pred(im, mask_gt, mask_pred)	Simple function for comparing oringal image and ground truth image
compute_pca(data_values[, n_components])	Computes PCA given input data
compute_tsne(data_values[, random_state])	Computes t-distributed stochastic neighbor embedding.
compute_umap(data_values, **kwargs)	param data_values: array of values to input into the reducer. Rows should be roi values,
image_show(image)	param image: image to show.
mask_to_outlines(mask[, im, binary_mask, debug])	Creates an outline of regions based on the input labels mask.

cell_analysis_tools.visualization.compare_images(title1, im1, title2, im2, suptitle=None, figsize=(10, 5), save_path=None) → None

Parameters:

• im1 (np.ndarray) – image 1.

• title1 (str) – title for image 1.

• im2 (np.ndarray) – image 2.

• title2 (str) – title for image 2.

• figsize (TYPE, optional) – size of figure. The default is (10, 5).

Return type:

None.

cell_analysis_tools.visualization.compare_orig_mask_gt_pred(im: ndarray, mask_gt: ndarray, mask_pred: ndarray, title: str = '') → None

Simple function for comparing oringal image and ground truth image

Parameters:

• im (np.ndarray) – original image.

• mask_gt (np.ndarray) – ground truth image.

• mask_pred (np.ndarray) – predicted mask.

• title (str, optional) – title of the plot, usually the origina filename. The default is “”.

Returns:

function only just plots data.

Return type:

None

cell_analysis_tools.visualization.compute_pca(data_values: ndarray, n_components: int = 2, **kwargs) → DataFrame

Computes PCA given input data

Parameters:

• data_values (np.ndarray) – array of values to input into the reducer. Rows should be roi values, columns are features

• n_components (int, optional) – number of components for the PCA. The default is 2.

Returns:

• df_pca (pd.DataFrame) – Dataframe containing the components of the PCA.

• pca (Object) – PCA reducer object.

cell_analysis_tools.visualization.compute_tsne(data_values: ndarray, random_state: int = 0, **kwargs) → ndarray

Computes t-distributed stochastic neighbor embedding.

Parameters:

• data_values (np.ndarray) – array of values to input into the reducer. Rows should be roi values, columns are features.

• random_state (int, optional) – allows consistent initialization of reducer. The default is 0.

Returns:

• df_tsne (pd.DataFrame) – Dataframe containing the embeddings of tsne.

• tsne (Object) – tsne reducer object.

cell_analysis_tools.visualization.compute_umap(data_values: ndarray, **kwargs) → DataFrame

Parameters:

• data_values (np.ndarray) – array of values to input into the reducer. Rows should be roi values, columns are features

• random_state (int, optional) – allows consistent initialization of reducer. The default is 0.

• **kwargs (dict) – additional parameters can be passed in here for the reducer.

Returns:

• df_umap (pd.DataFrame) – DataFrame containing umap embeddings.

• reducer (Object) – umap reducer object.

• .. note:: – See the umap-learn documentation for default and additional parameters https://umap-learn.readthedocs.io/en/latest/api.html

cell_analysis_tools.visualization.image_show(image)

Parameters:

image (ndarray) – image to show.

Return type:

None.

cell_analysis_tools.visualization.mask_to_outlines(mask, im=None, binary_mask=False, debug=False)

Creates an outline of regions based on the input labels mask.

Parameters:

• mask (np.ndarray) – Labels mask to convert to outlines.

• im (np.ndarray, optional) – DESCRIPTION. The default is None.

• binary_mask (bool, optional) – Determine if returned array should be boolean. The default is False.

• debug (TYPE, optional) – Displays intermediate images/masks for debugging. The default is False.

Returns:

mask_outline – Array containing outlines of input labeled mask.

Return type:

np.ndarray