API

Import Harpy as:

import harpy as hp

IO

I/O.

io.macsima(path[, image_name, ...])	Read MACSima formatted dataset.
io.merscope(path[, to_coordinate_system, ...])	Read MERSCOPE data from Vizgen.
io.xenium(path[, to_coordinate_system, ...])	Read a 10X Genomics Xenium dataset into a SpatialData object.
io.visium(path[, dataset_id, counts_file, ...])	Read 10x Genomics Visium formatted dataset.
io.visium_hd(path[, dataset_id, ...])	Read 10x Genomics Visium HD formatted dataset.
io.read_transcripts(sdata, path_count_matrix)	Reads transcript information from a file with each row listing the x and y coordinates, along with the gene name.
io.read_resolve_transcripts(sdata, ...[, ...])	Reads and adds transcripts from Resolve Biosciences’ Molecular Cartography technology to a SpatialData object.
io.read_merscope_transcripts(sdata, ...[, ...])	Reads and adds merscope transcript information to a SpatialData object.
io.read_stereoseq_transcripts(sdata, ...[, ...])	Reads and adds Stereoseq transcript information to a SpatialData object.
io.convert_to_zarr_2(sdata, output[, overwrite])	Convert a backed Zarr v3 SpatialData object into a Zarr v2 SpatialData store.

Image

Operations on image and labels elements.

im.add_image(sdata, arr, output_image_name)	Add an image element to a SpatialData object.
im.add_labels(sdata, arr, output_labels_name)	Add a labels element to a SpatialData object.
im.get_dataarray(sdata, element_name[, scale])	Retrieve the highest-resolution xarray.DataArray from an element in sdata.images or sdata.labels.
im.map_image(sdata, image_name, ...[, ...])	Apply a specified function to an image element of a SpatialData object.
im.tiling_correction(sdata[, image_name, ...])	Function corrects for the tiling effect that occurs in some image data (e.g. resolve data).
im.enhance_contrast(sdata[, image_name, ...])	Enhance the contrast of an image in a SpatialData object.
im.normalize(sdata, image_name, ...[, ...])	Normalize the intensity of an image element in a SpatialData object using specified percentiles.
im.min_max_filtering(sdata[, image_name, ...])	Apply min max filtering to an image in a SpatialData object using dask (using dask_image.ndfilters.maximum_filter and dask_image.ndfilters.minimum_filter).
im.gaussian_filtering(sdata[, image_name, ...])	Apply Gaussian filtering to an image in a SpatialData object using dask.
im.transcript_density(sdata[, image_name, ...])	Calculate the transcript density using a Gaussian filter and add it to the provided spatialdata.SpatialData object.
im.combine(sdata, image_name, output_image_name)	Combines specific channels within an image element of a SpatialData object.
im.segment(sdata, image_name[, model, ...])	Segment images using a provided model and add segmentation results (labels element and shapes element) to the SpatialData object.
im.segment_points(sdata, labels_name, ...[, ...])	Segment images using a points_name and a prior (labels_name) and add segmentation results (labels element and shapes element) to the SpatialData object.
im.cellpose_callable(img[, batch_size, ...])	Perform cell segmentation using the Cellpose model.
im.instanseg_callable(img[, device, ...])	Perform segmentation using instanseg.
im.baysor_callable(img, df, name_x, name_y, ...)	Perform cell segmentation using the Baysor algorithm.
im.add_grid_labels(sdata, shape, size, ...)	Adds a grid-based labels element to the SpatialData object using either a hexagonal or square grid.
im.expand_labels(sdata, labels_name[, ...])	Expand cells in the labels element labels_name using skimage.segmentation.expand_labels.
im.align_labels(sdata, labels_name_1, ...[, ...])	Align two labels elements.
im.map_labels(sdata, func, labels_name[, ...])	Apply a specified function to a labels element in a SpatialData object.
im.filter_labels(sdata, labels_name[, ...])	Filter labels in a labels element by global object size.
im.merge_labels(sdata, ...[, threshold, ...])	Merge two labels elements using a global object-level overlap rule.
im.match_labels_to_reference(sdata, ...[, ...])	Match source labels to reference labels based on an overlap score.
im.rasterize(sdata, shapes_name, ...[, ...])	Given a shapes element in a SpatialData object, corresponding masks are created, and added as a labels element to the SpatialData object.
im.pixel_clustering_preprocess(sdata, ...[, ...])	Preprocess image elements specified in image_name.
im.flowsom(sdata, image_name, ...[, ...])	Applies flowsom clustering on image element(s) of a SpatialData object.

Shape

Operations on shapes (polygons) elements.

sh.vectorize(sdata, labels_name, ...[, ...])	Vectorize a labels element.
sh.add_shapes(sdata, input, output_shapes_name)	Add a shapes element to a SpatialData object.
sh.filter_shapes(sdata, table_name, ...)	Filter shapes in a SpatialData object.
sh.create_voronoi_boundaries(sdata[, ...])	Create Voronoi boundaries from the shapes element of the provided SpatialData object.

Table

Operations on table (AnnData object) elements.

tb.add_table(sdata, adata, ...[, ...])	Add an AnnData object as a table element to a SpatialData object.
tb.allocate(sdata, labels_name[, ...])	Allocates transcripts to instances via provided labels_name and points_name and returns updated SpatialData object with a table element (sdata.tables[output_table_name]) holding the anndata.AnnData object with transcript counts.
tb.bin_counts(sdata, table_name, ...[, ...])	Bins gene counts from barcodes to cells or regions defined in labels_name and returns an updated SpatialData object with a table element (sdata.tables[output_table_name]) holding an AnnData object with the binned counts per cell or region.
tb.allocate_intensity(sdata[, image_name, ...])	Allocates intensity values from a specified image element to corresponding cells in a SpatialData object and returns an updated SpatialData object augmented with a table element (sdata.tables[output_table_name]) AnnData object with intensity values for each cell and each (specified) channel.
tb.add_regionprops(sdata, labels_name, ...)	Calculates region property features from the specified labels element, and adds the results to the AnnData object that annotates the labels element.
tb.add_feature_matrix(sdata, labels_name, ...)	Compute per-instance feature matrices from labels and optional image data.
tb.extract_instances(sdata, image_name, ...)	Extract per-label instance windows from image_name/labels_name of size diameter in y and x using dask.array.map_overlap() and dask.array.map_blocks().
tb.ZarrIterableInstances(zarr_path, instance_ids)	Chunk-wise iterable dataset that:
tb.ZarrDataLoader(*args[, start_epoch])	DataLoader that increments epoch and forwards it to epoch-aware datasets.
tb.featurize(sdata, image_name, labels_name, ...)	Extract per-instance feature vectors from image_name and labels_name using a user-provided embedding model.
tb.preprocess_transcriptomics(sdata, ...[, ...])	Preprocess a table (AnnData) attribute of a SpatialData object for transcriptomics data.
tb.preprocess_proteomics(sdata, labels_name, ...)	Preprocess a table (AnnData) attribute of a SpatialData object for proteomics data.
tb.filter_on_size(sdata, labels_name, ...[, ...])	Returns the updated SpatialData object.
tb.leiden(sdata, labels_name, table_name, ...)	Applies leiden clustering on the table_name of the SpatialData object with optional UMAP calculation and gene ranking.
tb.kmeans(sdata, labels_name, table_name, ...)	Applies KMeans clustering on the table_name of the SpatialData object with optional UMAP calculation and gene ranking.
tb.score_genes(sdata, labels_name, ...[, ...])	The function loads marker genes from a CSV file and scores cells for each cell type using those markers using scanpy's score_genes() function.
tb.score_genes_iter(sdata, labels_name, ...)	Iterative annotation algorithm.
tb.correct_marker_genes(sdata, labels_name, ...)	Correct celltype expression in sdata.tables[table_name] using celltype_correction_dict.
tb.cluster_cleanliness(sdata, labels_name, ...)	Re-calculates annotations, potentially following corrections to the list of celltypes, or after a manual update of the assigned scores per cell type via e.g. correct_marker_genes.
tb.nhood_enrichment(sdata, labels_name, ...)	Calculate the nhood enrichment using squidpy via squidpy.gr.spatial_neighbors() and squidpy.gr.nhood_enrichment().
tb.nhood_kmeans(sdata, table_name, ...[, ...])	Cluster cells (instances) based on neighborhood cell-type composition using KMeans.
tb.cluster_intensity(sdata, table_name, ...)	Calculates weighted (by instance size) average intensity per cluster.
tb.cluster_intensity_SOM(sdata, mapping, ...)	Calculates average intensity of each channel in image_name per SOM cluster as available in the labels_name, and saves it as a table element in sdata as output_table_name.
tb.spatial_pixel_neighbors(sdata, labels_name)	Computes spatial pixel neighbors and performs neighborhood enrichment analysis.
tb.cell_clustering_preprocess(sdata, ...[, ...])	Preprocesses spatial data for cell clustering.
tb.flowsom(sdata, cells_labels_name, ...[, ...])	Run FlowSOM cell clustering on pixel-cluster-derived cell features.
tb.weighted_channel_expression(sdata, ...[, ...])	Calculation of weighted channel expression in the context of cell clustering.

Points

Operations on points (Dask DataFrame object) elements.

pt.add_points(sdata, ddf, ...[, ...])

Add a points element to a SpatialData object.

Externals

External integrations.

externals.ilastik.run_object_classification(...)

Run ilastik headless object classification and add predicted labels to a table element.

Quality Control

Quality control functions.

image_histogram(sdata, image_name, channel, bins)	Generate and visualize a histogram for a specified image channel within an image of a SpatialData object.
segmentation_coverage(sdata, labels_name[, ...])	Calculate coverage statistics for a segmentation labels element.
segmentation_histogram(sdata, labels_name[, ...])	Plot a histogram of segmented instance sizes for a labels element.
analyse_genes_left_out(sdata, labels_name, ...)	Analyse and visualize the proportion of genes that could not be assigned to an instance during allocation step.
metric_histogram(sdata, table_name[, ...])	Plot a QC metric histogram for an AnnData table.
metrics_histogram(sdata, table_name[, ...])	Plot a standard panel of QC metric histograms for an AnnData table.
obs_scatter(sdata, table_name[, ...])	Plot the relationship between two observation-level columns.

Plotting

Plotting functions.

General plots

plot_sdata(sdata, image_name[, channel, ...])	Light wrapper around spatialdata-plot to plot a SpatialData object.
plot_sdata_genes(sdata, points_name[, ...])	Light wrapper around spatialdata-plot to visualize gene expression from a SpatialData object.
plot_instance_density(sdata, table_name[, ...])	Plot an instance density heatmap from centroids stored in sdata.tables[table_name].obsm[spatial_key].

Proteomics plots

cluster_intensity_heatmap(sdata, table_name, ...)	Generate and visualize a heatmap of mean channel intensities per cluster for each channel.
pixel_clusters(sdata, labels_name[, crd, ...])	Visualize spatial distribution of pixel clusters based on labels in a SpatialData object, obtained using harpy.im.flowsom().
pixel_clusters_heatmap(sdata, table_name[, ...])	Generate and visualize a heatmap of mean channel intensities for clusters or metaclusters.
snr_ratio(sdata[, ax, loglog, color])	Plot the signal to noise ratio.
group_snr_ratio(sdata, groupby[, ax, ...])	Plot the signal to noise ratio.
snr_clustermap(sdata[, signal_threshold, ...])
signal_clustermap(sdata[, signal_threshold, ...])
clustermap(*args, **kwargs)

Transcriptomics plots

plot_transcript_density(sdata, bin_size, ...)

Plot a transcript density heatmap from a SpatialData object.

Utils

Utility functions.

utils.RasterAggregator(mask_dask_array[, ...])	Helper class to calulate aggregated 'sum', 'mean', 'var', 'kurtosis', 'skew', 'area', 'min', 'max' and 'center of mass' of image and labels using Dask.
utils.Featurizer(mask_dask_array[, ...])	Helper class to featurize images and labels using Dask.
utils.kronos_embedding(array, ...[, ...])	Compute KRONOS embeddings for multi-channel instance windows using a pre-trained vision transformer.
utils.bounding_box_query(sdata, labels_name, ...)	Query the labels elements of a SpatialData object and the corresponding instances it annotates in sdata.tables via a bounding box query.

Datasets

Dataset loaders.

datasets.cluster_blobs([shape, ...])	Differs from spatialdata.datasets.make_blobs in that it generates cells with multiple image channels and known ground truth cell types.
datasets.multisample_blobs([n_samples, prefix])	Multisample blobs.
datasets.pixie_example()	Example pixie dataset, loaded from s3 bucket.
datasets.macsima_example()	Example proteomics dataset generated using the MACSima platform.
datasets.macsima_colorectal_carcinoma([...])	Load the Colorectal Carcinoma MACSima dataset as a spatialdata.SpatialData object.
datasets.macsima_colorectal_carcinoma_course(...)	Colorectal carcinoma MACSima course dataset.
datasets.macsima_tonsil([filter_regex, path])	Tonsil proteomics dataset generated using the MACSima platform
datasets.codex_example([path])	Example annotated codex dataset (cHL maps dataset), Shaban, M.
datasets.mibi_example()	Example proteomics dataset
datasets.vectra_example([path])	Example proteomics dataset LuCa-7color_[13860,52919]_1x1 from Perkin Elmer
datasets.resolve_example([path])	Example transcriptomics dataset.
datasets.merscope_mouse_liver([output, ...])	Example transcriptomics dataset
datasets.xenium_human_lung_cancer([output, path])	Example transcriptomics dataset
datasets.xenium_human_ovarian_cancer([...])	Example transcriptomics dataset
datasets.xenium_human_ovarian_cancer_course(...)	Human ovarian cancer Xenium course dataset.
datasets.visium_hd_example([bin_size, ...])	Example transcriptomics dataset
datasets.visium_hd_example_custom_binning([path])	Example transcriptomics dataset
datasets.get_registry([path])	Get the Pooch registry
datasets.get_spatialdata_registry([path])	Get the Pooch SpatialData registry