Add Plotting Code

pyproject.toml

@@ -54,6 +54,7 @@ test = ["pytest", "pytest-mock", "tox", "coverage"]
 preprocess = "nuc2seg.cli.preprocess:main"
 train = "nuc2seg.cli.train:main"
 predict = "nuc2seg.cli.predict:main"
+plot_predictions = "nuc2seg.cli.plot_predictions:main"
 
 [build-system]
 requires = ["setuptools>=43.0.0", "wheel"]

src/nuc2seg/cli/plot_predictions.py

@@ -0,0 +1,100 @@
+import argparse
+import logging
+import os.path
+import tqdm
+
+from nuc2seg import log_config
+from nuc2seg.data import Nuc2SegDataset, TiledDataset, ModelPredictions, generate_tiles
+from nuc2seg.plotting import plot_model_predictions
+
+logger = logging.getLogger(__name__)
+
+
+def get_parser():
+    parser = argparse.ArgumentParser(
+        description="Evaluate a UNet model on preprocessed data."
+    )
+    log_config.add_logging_args(parser)
+    parser.add_argument(
+        "--predictions",
+        help="Model prediction output in h5 format.",
+        type=str,
+        required=True,
+    )
+    parser.add_argument(
+        "--dataset",
+        help="Path to dataset in h5 format.",
+        type=str,
+        required=True,
+    )
+    parser.add_argument(
+        "--tile-height",
+        help="Height of the tiles.",
+        type=int,
+        default=64,
+    )
+    parser.add_argument(
+        "--tile-width",
+        help="Width of the tiles.",
+        type=int,
+        default=64,
+    )
+    parser.add_argument(
+        "--overlap-percentage",
+        help="What percent of each tile dimension overlaps with the next tile.",
+        type=float,
+        default=0.25,
+    )
+    parser.add_argument(
+        "--output-dir",
+        help="Output directory for plots.",
+        type=str,
+        required=True,
+    )
+    return parser
+
+
+def get_args():
+    parser = get_parser()
+
+    args = parser.parse_args()
+
+    return args
+
+
+def main():
+    args = get_args()
+
+    log_config.configure_logging(args)
+
+    logger.info(f"Loading dataset from {args.dataset}")
+
+    ds = Nuc2SegDataset.load_h5(args.dataset)
+
+    tiled_dataset = TiledDataset(
+        ds,
+        tile_height=args.tile_height,
+        tile_width=args.tile_width,
+        tile_overlap=args.overlap_percentage,
+    )
+    predictions = ModelPredictions.load_h5(args.predictions)
+
+    tile_generator = generate_tiles(
+        tiler=tiled_dataset.tiler,
+        x_extent=ds.x_extent_pixels,
+        y_extent=ds.y_extent_pixels,
+        tile_size=(args.tile_width, args.tile_height),
+        overlap_fraction=args.overlap_percentage,
+    )
+
+    os.makedirs(args.output_dir, exist_ok=True)
+
+    for bbox in tqdm.tqdm(tile_generator, total=len(tiled_dataset), unit="plot"):
+        plot_model_predictions(
+            dataset=ds,
+            model_predictions=predictions,
+            bbox=bbox,
+            output_path=os.path.join(
+                args.output_dir, "_".join([str(x) for x in bbox]) + ".pdf"
+            ),
+        )

src/nuc2seg/plotting.py

@@ -1,6 +1,9 @@
 import geopandas
 from shapely import box
 from matplotlib import pyplot as plt
+from nuc2seg.data import Nuc2SegDataset, ModelPredictions
+import numpy as np
+import math
 
 
 def plot_tiling(bboxes, output_path):
@@ -14,5 +17,68 @@ def plot_tiling(bboxes, output_path):
     fig.savefig(output_path)
 
 
-def plot_model_predictions():
-    pass
+def plot_labels(ax, dataset: Nuc2SegDataset, bbox=None):
+    label_plot = dataset.labels.copy()
+    transcripts = dataset.transcripts.copy()
+    label_plot[label_plot >= 1] = 5
+    label_plot[label_plot == -1] = 2
+
+    if bbox is not None:
+        label_plot = label_plot[bbox[0] : bbox[2], bbox[1] : bbox[3]]
+        mask_x = (transcripts[:, 0] >= bbox[0]) & (transcripts[:, 0] < bbox[2])
+        mask_y = (transcripts[:, 1] >= bbox[1]) & (transcripts[:, 1] < bbox[3])
+        mask = mask_x & mask_y
+        transcripts = transcripts[mask]
+
+        transcripts[:, 0] = transcripts[:, 0] - bbox[0]
+        transcripts[:, 1] = transcripts[:, 1] - bbox[1]
+
+    ax.set_title("Labels and transcripts")
+    ax.imshow(label_plot, cmap="copper", interpolation="none")
+
+    ax.scatter(
+        transcripts[:, 1], transcripts[:, 0], color="red", zorder=100, s=0.1, alpha=0.3
+    )
+
+
+def plot_angles(ax, predictions: ModelPredictions, skip_factor=1, bbox=None):
+    angles = predictions.angles
+
+    if bbox is not None:
+        angles = angles[bbox[0] : bbox[2], bbox[1] : bbox[3]]
+
+    skip = (slice(None, None, skip_factor), slice(None, None, skip_factor))
+
+    ax.set_title("Predicted angles")
+
+    U = 0.5 * np.cos(angles)
+    V = 0.5 * np.sin(angles)
+    Y, X = np.mgrid[0 : U.shape[0], 0 : U.shape[1]]
+
+    ax.quiver(X[skip], Y[skip], U[skip], V[skip], color="black", headwidth=2)
+
+
+def plot_foreground(ax, predictions: ModelPredictions, bbox=None):
+    foreground = predictions.foreground
+
+    if bbox is not None:
+        foreground = foreground[bbox[0] : bbox[2], bbox[1] : bbox[3]]
+
+    ax.imshow(foreground, vmin=0, vmax=1, cmap="coolwarm", interpolation="none")
+    ax.set_title("Predicted foreground")
+
+
+def plot_model_predictions(
+    dataset: Nuc2SegDataset,
+    model_predictions: ModelPredictions,
+    output_path=None,
+    bbox=None,
+):
+    fig, ax = plt.subplots(figsize=(10, 10), nrows=3, dpi=1000)
+
+    plot_labels(ax[0], dataset, bbox=bbox)
+    plot_angles(ax[1], model_predictions, bbox=bbox)
+    plot_foreground(ax[2], model_predictions, bbox=bbox)
+
+    fig.tight_layout()
+    fig.savefig(output_path)