dsa-helpers 1.0.7.dev0__tar.gz → 1.0.8.dev1__tar.gz

{dsa_helpers-1.0.7.dev0/dsa_helpers.egg-info → dsa_helpers-1.0.8.dev1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: dsa_helpers
-Version: 1.0.7.dev0
+Version: 1.0.8.dev1
 Summary: Utility functions for working with the DSA girder client.
 Author-email: "Juan C. Vizcarra" <jvizcar@emory.edu>
 Maintainer-email: "Juan C. Vizcarra" <jvizcar@emory.edu>

dsa_helpers-1.0.8.dev1/dsa_helpers/gpd_utils.py

@@ -0,0 +1,172 @@
+"""
+Utility functions for working with GeoPandas.
+
+"""
+
+import geopandas as gpd
+from shapely.geometry import Polygon
+
+import matplotlib.patches as mpatches
+import matplotlib.pyplot as plt
+
+
+def count_polygon_points(polygon: Polygon) -> int:
+    """
+    Count the total number of points in a Shapely polygon, including
+    exterior and interior rings.
+
+    Args:
+        polygon (shapely.geometry.Polygon): A Shapely polygon object.
+
+    Returns:
+        int: The total number of points in the polygon.
+
+    """
+    # Count exterior points.
+    exterior_points = len(polygon.exterior.coords)
+
+    # Count interior points (holes).
+    interior_points = sum(
+        len(interior.coords) for interior in polygon.interiors
+    )
+
+    # Total points.
+    return exterior_points + interior_points
+
+
+def total_points_in_gdf(gdf: gpd.GeoDataFrame) -> int:
+    """
+    Calculate the total number of points in all polygons in a GeoDataFrame.
+
+    Args:
+        gdf (geopandas.GeoDataFrame): GeoDataFrame containing polygons.
+
+    Returns:
+        int: The total number of points in all polygons.
+
+    """
+    return int(gdf.geometry.apply(count_polygon_points).sum())
+
+
+def plot_gdf(
+    gdf: gpd.GeoDataFrame,
+    figsize: int | tuple[int, int] | None = None,
+    label2color: dict | None = None,
+):
+    """Plot a geopandas dataframe.
+
+    Args:
+        gdf (geopandas.GeoDataFrame): A GeoDataFrame containing
+            polygons. Must have a 'label' column.
+        figsize (int | tuple[int, int] | None): The size of the figure.
+            If None, the figure size will be (5, 5).
+        label2color (dict | None): A dictionary mapping labels to
+            colors. If None then a legend will not be included and the
+            colors will be random. Colors should be an accepted
+            matplotlib color.
+
+    """
+    if figsize is None:
+        figsize = (5, 5)
+    elif isinstance(figsize, int):
+        figsize = (figsize, figsize)
+
+    gdf = gdf.copy()
+
+    # Blow up any multi-polygons.
+    gdf = gdf.explode(index_parts=False).reset_index(drop=True)
+
+    total_points = total_points_in_gdf(gdf)
+
+    if label2color is not None:
+        gdf["color"] = gdf["label"].map(label2color)
+
+        # Remove rows where color is nan.
+        gdf = gdf.dropna(subset=["color"])
+
+    # # Create the plot
+    ax = plt.subplots(figsize=figsize)[1]
+    gdf.plot(color=gdf["color"], ax=ax)
+
+    if label2color is not None:
+        # Add legend.
+        handles = [
+            mpatches.Patch(color=color, label=key)
+            for key, color in label2color.items()
+        ]
+
+        ax.legend(
+            handles=handles,
+            title="Legend",
+            loc="upper left",
+            bbox_to_anchor=(
+                1.02,
+                1,
+            ),  # x > 1 pushes it to the right, y = 1 keeps it top-aligned
+            borderaxespad=0,
+        )
+    plt.title(f"Total points: {total_points:,}")
+    plt.show()
+
+
+def remove_gdf_overlaps(
+    gdf: gpd.GeoDataFrame,
+    columns: str | list[str],
+    ascending: bool | list[bool] = True,
+    is_notebook: bool = False,
+) -> gpd.GeoDataFrame:
+    """Remove overlaps from a GeoDataFrame.
+
+    Args:
+        gdf (geopandas.GeoDataFrame): The GeoDataFrame to remove
+            overlaps from.
+        columns (str | list[str]): The columns to sort dataframe by. The
+            order determines the order of removal overlapping polygon
+            regions.
+        ascending (bool): Whether to remove overlaps in ascending order.
+        is_notebook (bool): Switch which tqdm bar to use. Default is
+            False, which uses the terminal progress bar.
+
+    Returns:
+        gpd.GeoDataFrame: The GeoDataFrame with overlaps removed.
+
+    """
+    if isinstance(columns, str):
+        columns = [columns]
+
+    if isinstance(ascending, bool):
+        ascending = [ascending] * len(columns)
+
+    # Remove multi-polygons.
+    gdf = gdf.explode(index_parts=False).reset_index(drop=True)
+
+    # Sort the GeoDataFrame by the columns and ascending order.
+    gdf = gdf.sort_values(by=columns, ascending=ascending).reset_index(
+        drop=True
+    )
+
+    if is_notebook:
+        from tqdm.notebook import tqdm
+    else:
+        from tqdm import tqdm
+
+    # Loop through each row in the GeoDataFrame.
+    for i, row in tqdm(gdf.iterrows(), total=len(gdf)):
+        # Check overlap with all following rows.
+        other_gdf = gdf[i + 1 :]
+
+        if len(other_gdf):
+            # Get the rows that overlap.
+            overlap = row.geometry.overlaps(other_gdf.geometry)
+
+            # Get the rows that overlap.
+            other_rows = other_gdf[overlap]
+
+            other_gdf.loc[:, "geometry"] = other_gdf.geometry.difference(
+                row.geometry
+            )
+
+    # This could have created multiple polygons, explode them.
+    gdf = gdf.explode(index_parts=False).reset_index(drop=True)
+
+    return gdf

dsa_helpers-1.0.8.dev1/dsa_helpers/utils.py

@@ -0,0 +1,150 @@
+"""Utility functions for DSA Helpers.
+
+This module provides various miscellaneous utility functions that are
+not grouped into their own modules.
+"""
+
+from shapely.geometry import Polygon
+import numpy as np
+import cv2 as cv
+
+
+def remove_small_holes(
+    polygon: Polygon, hole_area_threshold: float
+) -> Polygon:
+    """Remove small holes from a shapely polygon.
+
+    Args:
+        polygon (shapely.geometry.Polygon): Polygon to remove holes
+            from.
+        hole_area_threshold (float): Minimum area of a hole to keep it.
+
+    Returns:
+        shapely.geometry.Polygon: Polygon with small holes removed.
+
+    """
+    if not polygon.interiors:  # if there are no holes, return as is
+        return polygon
+
+    # Filter out small holes
+    new_holes = [
+        hole
+        for hole in polygon.interiors
+        if Polygon(hole).area > hole_area_threshold
+    ]
+
+    # Create a new polygon with only large holes
+    return Polygon(polygon.exterior, new_holes)
+
+
+def convert_to_json_serializable(
+    data: int | float | str | list | dict,
+) -> int | float | str | list | dict:
+    """Convert a list, integer, float, or dictionary into a JSON
+    serializable version of the object. Uses recursion to make sure the
+    entire input data structure is converted to a JSON serializable
+    version.
+
+    Args:
+        data (int | float | str | list | dict): Data to convert to
+            JSON serializable.
+
+    Returns:
+        int | float | str | list | dict: JSON serializable version of
+            the input data.
+
+    """
+    if isinstance(data, np.ndarray):
+        return data.tolist()
+    elif isinstance(data, np.integer):
+        return int(data)
+    elif isinstance(data, np.floating):
+        return float(data)
+    elif isinstance(data, dict):
+        return {
+            key: convert_to_json_serializable(value)
+            for key, value in data.items()
+        }
+    elif isinstance(data, list):
+        return [convert_to_json_serializable(item) for item in data]
+    return data
+
+
+def contours_to_polygons(binary_image: np.ndarray) -> list[Polygon]:
+    """Convert a binary image to a list of polygons.
+
+    Args:
+        binary_image (numpy.ndarray): Binary image.
+
+    Returns:
+        list[shapely.geometry.Polygon]: List of polygons.
+
+    """
+    contours, hierarchy = cv.findContours(
+        binary_image, cv.RETR_TREE, cv.CHAIN_APPROX_SIMPLE
+    )
+    hierarchy = hierarchy[0]  # shape: (n_contours, 4)
+
+    def contour_to_coords(contour):
+        return [(int(p[0][0]), int(p[0][1])) for p in contour]
+
+    polygons = []
+    used = set()
+
+    for i, (next_idx, prev_idx, child_idx, parent_idx) in enumerate(hierarchy):
+        if parent_idx != -1:
+            # Skip child contours for now (they'll be added as holes)
+            continue
+
+        # Outer contour
+        exterior = contour_to_coords(contours[i])
+
+        if len(exterior) < 4:
+            # Skip contours that are too small
+            used.add(i)
+            continue
+
+        holes = []
+
+        # Look for children (holes)
+        child = hierarchy[i][2]
+        while child != -1:
+            hole_coords = contour_to_coords(contours[child])
+            holes.append(hole_coords)
+            used.add(child)
+            # Check for nested children (e.g. box in a hole in a box)
+            grandchild = hierarchy[child][2]
+            if grandchild != -1:
+                # Treat grandchild as a new polygon later
+                pass
+            child = hierarchy[child][0]  # next sibling hole
+
+        used.add(i)
+        polygon = Polygon(exterior, holes)
+
+        if not polygon.is_valid:
+            # Attempt to fix the polygon
+            polygon = polygon.buffer(0)
+
+        if polygon.is_valid:
+            polygons.append(polygon)
+
+    # Add any unused outer contours (e.g. nested objects)
+    for i in range(len(contours)):
+        if i not in used:
+            coords = contour_to_coords(contours[i])
+
+            if len(coords) < 4:
+                # Skip contours that are too small
+                continue
+
+            poly = Polygon(coords)
+
+            if not poly.is_valid:
+                # Attempt to fix the polygon
+                poly = poly.buffer(0)
+
+            if poly.is_valid:
+                polygons.append(poly)
+
+    return polygons

{dsa_helpers-1.0.7.dev0 → dsa_helpers-1.0.8.dev1/dsa_helpers.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: dsa_helpers
-Version: 1.0.7.dev0
+Version: 1.0.8.dev1
 Summary: Utility functions for working with the DSA girder client.
 Author-email: "Juan C. Vizcarra" <jvizcar@emory.edu>
 Maintainer-email: "Juan C. Vizcarra" <jvizcar@emory.edu>

{dsa_helpers-1.0.7.dev0 → dsa_helpers-1.0.8.dev1}/dsa_helpers.egg-info/SOURCES.txt

@@ -10,6 +10,7 @@ docs/source/index.rst
 dsa_helpers/__init__.py
 dsa_helpers/cli_utils.py
 dsa_helpers/girder_utils.py
+dsa_helpers/gpd_utils.py
 dsa_helpers/image_utils.py
 dsa_helpers/imread.py
 dsa_helpers/imwrite.py

{dsa_helpers-1.0.7.dev0 → dsa_helpers-1.0.8.dev1}/pyproject.toml

@@ -35,7 +35,7 @@ classifiers = [
     "License :: OSI Approved :: MIT License",
 ]
 keywords = ["dsa", "girder", "girder_client"]
-version = "v1.0.7.dev0"
+version = "v1.0.8.dev1"
 
 [project.urls]
 Repository = "https://github.com/Gutman-Lab/dsa-python-utils"

dsa_helpers-1.0.7.dev0/dsa_helpers/utils.py

@@ -1,69 +0,0 @@
-"""Utility functions for DSA Helpers.
-
-This module provides various miscellaneous utility functions that are
-not grouped into their own modules.
-"""
-
-from shapely.geometry import Polygon
-import numpy as np
-
-
-def remove_small_holes(
-    polygon: Polygon, hole_area_threshold: float
-) -> Polygon:
-    """Remove small holes from a shapely polygon.
-
-    Args:
-        polygon (shapely.geometry.Polygon): Polygon to remove holes
-            from.
-        hole_area_threshold (float): Minimum area of a hole to keep it.
-
-    Returns:
-        shapely.geometry.Polygon: Polygon with small holes removed.
-
-    """
-    if not polygon.interiors:  # if there are no holes, return as is
-        return polygon
-
-    # Filter out small holes
-    new_holes = [
-        hole
-        for hole in polygon.interiors
-        if Polygon(hole).area > hole_area_threshold
-    ]
-
-    # Create a new polygon with only large holes
-    return Polygon(polygon.exterior, new_holes)
-
-
-def convert_to_json_serializable(
-    data: int | float | str | list | dict,
-) -> int | float | str | list | dict:
-    """Convert a list, integer, float, or dictionary into a JSON
-    serializable version of the object. Uses recursion to make sure the
-    entire input data structure is converted to a JSON serializable
-    version.
-
-    Args:
-        data (int | float | str | list | dict): Data to convert to
-            JSON serializable.
-
-    Returns:
-        int | float | str | list | dict: JSON serializable version of
-            the input data.
-
-    """
-    if isinstance(data, np.ndarray):
-        return data.tolist()
-    elif isinstance(data, np.integer):
-        return int(data)
-    elif isinstance(data, np.floating):
-        return float(data)
-    elif isinstance(data, dict):
-        return {
-            key: convert_to_json_serializable(value)
-            for key, value in data.items()
-        }
-    elif isinstance(data, list):
-        return [convert_to_json_serializable(item) for item in data]
-    return data