From f1189768c3ba2edcc998d05c498e3d00933e2e8e Mon Sep 17 00:00:00 2001
From: qh681248 <181246904+qh681248@users.noreply.github.com>
Date: Fri, 13 Dec 2024 12:49:18 +0000
Subject: [PATCH 1/8] feat: Initial implementation of video benchmark

#892
---
 benchmark/pounce_benchmark.py | 87 +++++++++++++++++++++++++++++++++++
 1 file changed, 87 insertions(+)
 create mode 100644 benchmark/pounce_benchmark.py

diff --git a/benchmark/pounce_benchmark.py b/benchmark/pounce_benchmark.py
new file mode 100644
index 00000000..ec0029af
--- /dev/null
+++ b/benchmark/pounce_benchmark.py
@@ -0,0 +1,87 @@
+# © Crown Copyright GCHQ
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""
+Benchmark performance of different coreset algorithms on frames of a video.
+
+The benchmarking process follows these steps:
+1. Load an input GIF and preprocess its frames.
+2. Reshape the frame data and apply UMAP for dimensionality reduction.
+3. Generate coresets using each algorithm and save the selected frames as GIFs.
+4. Print the time taken to generate each coreset.
+"""
+
+import os
+import time
+from pathlib import Path
+
+import imageio
+import numpy as np
+import umap
+from jax import random
+from mnist_benchmark import get_solver_name, initialise_solvers
+
+from coreax.data import Data
+
+
+def benchmark_coreset_algorithms(
+    in_path: Path = Path("../examples/data/pounce/pounce.gif"),
+    out_dir: Path = Path("pounce"),
+    coreset_size: int = 10,
+):
+    """
+    Benchmark coreset algorithms by processing a video GIF.
+
+    :param in_path: Path to the input GIF file.
+    :param out_dir: Directory to save the output GIFs for each coreset algorithm.
+    :param coreset_size: The size of the coreset.
+    """
+    base_dir = os.path.dirname(os.path.abspath(__file__))
+
+    # Ensure paths are absolute and output directory exists
+    in_path = Path(os.path.join(base_dir, in_path)).resolve()
+    out_dir = Path(os.path.join(base_dir, out_dir)).resolve()
+    out_dir.mkdir(parents=True, exist_ok=True)
+
+    # Load and preprocess video frames
+    _, *image_data = imageio.v2.mimread(in_path)
+    raw_data = np.asarray(image_data)
+    reshaped_data = raw_data.reshape(raw_data.shape[0], -1)
+    print(type(reshaped_data))
+
+    umap_model = umap.UMAP(densmap=True, n_components=25)
+    umap_data = umap_model.fit_transform(reshaped_data)
+
+    solvers = initialise_solvers(umap_data, random.PRNGKey(45))
+
+    for get_solver in solvers:
+        solver = get_solver(coreset_size)
+        solver_name = get_solver_name(solver)
+        data = Data(umap_data)
+
+        start_time = time.perf_counter()
+        coreset, _ = solver.reduce(data)
+        duration = time.perf_counter() - start_time
+
+        selected_indices = np.sort(np.asarray(coreset.unweighted_indices))
+
+        # Extract corresponding frames from original data and save GIF
+        coreset_frames = raw_data[selected_indices]
+        output_gif_path = out_dir / f"{solver_name}_coreset.gif"
+        imageio.mimsave(output_gif_path, coreset_frames)
+        print(f"Saved {solver_name} coreset GIF to {output_gif_path}")
+        print(f"time taken: {solver_name:<25} {duration:<30.4f}")
+
+
+if __name__ == "__main__":
+    benchmark_coreset_algorithms()

From bbfd64e1004b2ca32edaef603ea2b500972bdc53 Mon Sep 17 00:00:00 2001
From: qh681248 <181246904+qh681248@users.noreply.github.com>
Date: Fri, 13 Dec 2024 14:45:30 +0000
Subject: [PATCH 2/8] docs: Add video benchmarking to existing benchmarks #892

---
 CHANGELOG.md | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 09b1ab71..8b3c93d6 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -22,6 +22,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
     and KSD metrics) (https://github.com/gchq/coreax/pull/802)
   - David (extract pixel locations and values from an image and plot coresets side by
     side for visual benchmarking) (https://github.com/gchq/coreax/pull/880)
+  - Pounce (extract frames from a video and use coreset algorithms to select the best
+    frames) (https://github.com/gchq/coreax/issues/892)
 - `benchmark` dependency group for benchmarking dependencies.
   (https://github.com/gchq/coreax/pull/888)
 - Added a method `SquaredExponentialKernel.get_sqrt_kernel` which returns a square

From ac022c88cd422b686364eed51ac4e128bb676782 Mon Sep 17 00:00:00 2001
From: qh681248 <181246904+qh681248@users.noreply.github.com>
Date: Fri, 13 Dec 2024 15:29:52 +0000
Subject: [PATCH 3/8] chore: remove pylint comments to address pre-commit
 complains on remote #892

---
 benchmark/mnist_benchmark_coresets_only.py | 1 -
 examples/david_map_reduce_weighted.py      | 1 -
 2 files changed, 2 deletions(-)

diff --git a/benchmark/mnist_benchmark_coresets_only.py b/benchmark/mnist_benchmark_coresets_only.py
index 6fb5398e..f1aa0950 100644
--- a/benchmark/mnist_benchmark_coresets_only.py
+++ b/benchmark/mnist_benchmark_coresets_only.py
@@ -99,7 +99,6 @@ def main() -> None:
 
     # Run the experiment with 5 different random keys
     for i in range(5):
-        print(f"Run {i + 1} of 5:")
         key = jax.random.PRNGKey(i)
         solvers = initialise_solvers(train_data_umap, key)
         for getter in solvers:
diff --git a/examples/david_map_reduce_weighted.py b/examples/david_map_reduce_weighted.py
index d0acebf0..58133088 100644
--- a/examples/david_map_reduce_weighted.py
+++ b/examples/david_map_reduce_weighted.py
@@ -63,7 +63,6 @@
 
 # Examples are written to be easy to read, copy and paste by users, so we ignore the
 # pylint warnings raised that go against this approach
-# pylint: disable=no-member
 # pylint: disable=too-many-locals
 # pylint: disable=too-many-statements
 # pylint: disable=duplicate-code

From 7cbb306ef8cca5eaaf39c237c282aa43c7c6041c Mon Sep 17 00:00:00 2001
From: qh681248 <181246904+qh681248@users.noreply.github.com>
Date: Tue, 17 Dec 2024 12:47:34 +0000
Subject: [PATCH 4/8] chore: Set regularise to False for stein_solver as it
 does not behave properly for high dimensional data with regularise enabled
 and pounce_benchmark does not use MapReduce #892

---
 benchmark/mnist_benchmark.py  | 4 +++-
 benchmark/pounce_benchmark.py | 8 ++++++--
 2 files changed, 9 insertions(+), 3 deletions(-)

diff --git a/benchmark/mnist_benchmark.py b/benchmark/mnist_benchmark.py
index bf0a49f2..94c9b483 100644
--- a/benchmark/mnist_benchmark.py
+++ b/benchmark/mnist_benchmark.py
@@ -474,7 +474,9 @@ def _get_stein_solver(_size: int) -> MapReduce:
             train_data_umap[idx]
         )
         stein_kernel = SteinKernel(kernel, score_function)
-        stein_solver = SteinThinning(coreset_size=_size, kernel=stein_kernel)
+        stein_solver = SteinThinning(
+            coreset_size=_size, kernel=stein_kernel, regularise=False
+        )
         return MapReduce(stein_solver, leaf_size=3 * _size)
 
     def _get_random_solver(_size: int) -> RandomSample:
diff --git a/benchmark/pounce_benchmark.py b/benchmark/pounce_benchmark.py
index ec0029af..1d3d8c77 100644
--- a/benchmark/pounce_benchmark.py
+++ b/benchmark/pounce_benchmark.py
@@ -63,7 +63,11 @@ def benchmark_coreset_algorithms(
     umap_data = umap_model.fit_transform(reshaped_data)
 
     solvers = initialise_solvers(umap_data, random.PRNGKey(45))
-
+    # There is no need to use MapReduce as the data-size is small
+    solvers = [
+        solver.base_solver if solver.__class__.__name__ == "MapReduce" else solver
+        for solver in solvers
+    ]
     for get_solver in solvers:
         solver = get_solver(coreset_size)
         solver_name = get_solver_name(solver)
@@ -78,7 +82,7 @@ def benchmark_coreset_algorithms(
         # Extract corresponding frames from original data and save GIF
         coreset_frames = raw_data[selected_indices]
         output_gif_path = out_dir / f"{solver_name}_coreset.gif"
-        imageio.mimsave(output_gif_path, coreset_frames)
+        imageio.mimsave(output_gif_path, coreset_frames, loop=0)
         print(f"Saved {solver_name} coreset GIF to {output_gif_path}")
         print(f"time taken: {solver_name:<25} {duration:<30.4f}")
 

From c8e08edf67434e5e940b62a3dc85fb36b2ae2266 Mon Sep 17 00:00:00 2001
From: qh681248 <181246904+qh681248@users.noreply.github.com>
Date: Tue, 17 Dec 2024 14:47:26 +0000
Subject: [PATCH 5/8] chore: Add pylint disable for code repetition. #892

---
 tests/performance/compare.py | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/tests/performance/compare.py b/tests/performance/compare.py
index e507e3d9..d9113836 100644
--- a/tests/performance/compare.py
+++ b/tests/performance/compare.py
@@ -90,6 +90,7 @@ def parse_args() -> Tuple[Path, Path, str, Path]:
     )
 
 
+# pylint: disable=R0801
 def date_from_filename(path: Path) -> Optional[Tuple[datetime.datetime, str]]:
     """
     Extract the date from a performance data file name.
@@ -129,6 +130,7 @@ def date_from_filename(path: Path) -> Optional[Tuple[datetime.datetime, str]]:
     ), git_hash
 
 
+# pylint: disable=R0801
 def get_most_recent_historic_data(
     reference_directory: Path,
 ) -> FullPerformanceData:

From 2f7570d781f25d112c91fdffec6829a698d2e64d Mon Sep 17 00:00:00 2001
From: qh681248 <181246904+qh681248@users.noreply.github.com>
Date: Thu, 19 Dec 2024 10:43:41 +0000
Subject: [PATCH 6/8] chore: push with regularise set to False to run MNIST
 benchmark again on the GPU instance.

---
 benchmark/mnist_benchmark.py | 14 +++++++++++---
 1 file changed, 11 insertions(+), 3 deletions(-)

diff --git a/benchmark/mnist_benchmark.py b/benchmark/mnist_benchmark.py
index 94c9b483..8b622c30 100644
--- a/benchmark/mnist_benchmark.py
+++ b/benchmark/mnist_benchmark.py
@@ -111,12 +111,19 @@ def compute_metrics(logits: jnp.ndarray, labels: jnp.ndarray) -> dict[str, jnp.n
 
 
 class MLP(nn.Module):
-    """Multi-layer perceptron with optional batch normalization and dropout."""
+    """
+    Multi-layer perceptron with optional batch normalisation and dropout.
+
+    :param hidden_size: Number of units in the hidden layer.
+    :param output_size: Number of output units.
+    :param use_batchnorm: Whether to apply batch norm.
+    :param dropout_rate: Dropout rate to use during training.
+    """
 
     hidden_size: int
     output_size: int = 10
     use_batchnorm: bool = True
-    dropout_rate: float = 0.5
+    dropout_rate: float = 0.2
 
     @nn.compact
     def __call__(self, x: jnp.ndarray, training: bool = True) -> jnp.ndarray:
@@ -128,10 +135,11 @@ def __call__(self, x: jnp.ndarray, training: bool = True) -> jnp.ndarray:
         :return: Output logits of the network.
         """
         x = nn.Dense(self.hidden_size)(x)
+        if training:
+            x = nn.Dropout(rate=self.dropout_rate, deterministic=False)(x)
         if self.use_batchnorm:
             x = nn.BatchNorm(use_running_average=not training)(x)
         x = nn.relu(x)
-        x = nn.Dropout(rate=self.dropout_rate, deterministic=not training)(x)
         x = nn.Dense(self.output_size)(x)
         return x
 

From 2e96257a3ba8cd390fe5a8c8663d91366cbd411c Mon Sep 17 00:00:00 2001
From: qh681248 <181246904+qh681248@users.noreply.github.com>
Date: Thu, 19 Dec 2024 11:24:01 +0000
Subject: [PATCH 7/8] Rerun MNIST benchmark on GPU machine.

---
 benchmark/mnist_benchmark_results.json | 482 ++++++++++++-------------
 1 file changed, 241 insertions(+), 241 deletions(-)

diff --git a/benchmark/mnist_benchmark_results.json b/benchmark/mnist_benchmark_results.json
index de81c888..456c1dd0 100644
--- a/benchmark/mnist_benchmark_results.json
+++ b/benchmark/mnist_benchmark_results.json
@@ -2,537 +2,537 @@
   "RandomSample": {
     "25": {
       "0": {
-        "accuracy": 0.45518216490745544,
-        "time_taken": 36.396069547000025
+        "accuracy": 0.47499004006385803,
+        "time_taken": 21.129530332
       },
       "1": {
-        "accuracy": 0.510104238986969,
-        "time_taken": 39.19832314600001
+        "accuracy": 0.5094035863876343,
+        "time_taken": 21.620098648999942
       },
       "2": {
-        "accuracy": 0.43347349762916565,
-        "time_taken": 36.08024689900003
+        "accuracy": 0.4429771900177002,
+        "time_taken": 23.587054420000072
       },
       "3": {
-        "accuracy": 0.5120047926902771,
-        "time_taken": 25.586340088000043
+        "accuracy": 0.5266107320785522,
+        "time_taken": 20.613561648000086
       },
       "4": {
-        "accuracy": 0.4719884693622589,
-        "time_taken": 49.6169613750003
+        "accuracy": 0.4708881676197052,
+        "time_taken": 23.3405117719999
       }
     },
     "50": {
       "0": {
-        "accuracy": 0.6087996363639832,
-        "time_taken": 12.642228439999997
+        "accuracy": 0.6092996001243591,
+        "time_taken": 7.480583503999981
       },
       "1": {
-        "accuracy": 0.6126995086669922,
-        "time_taken": 19.304289521999976
+        "accuracy": 0.6098998785018921,
+        "time_taken": 9.492431753999995
       },
       "2": {
-        "accuracy": 0.5459998846054077,
-        "time_taken": 18.504275409
+        "accuracy": 0.5316997766494751,
+        "time_taken": 7.016472408000027
       },
       "3": {
-        "accuracy": 0.6301995515823364,
-        "time_taken": 14.654397104999816
+        "accuracy": 0.6038997769355774,
+        "time_taken": 7.433440193000024
       },
       "4": {
-        "accuracy": 0.6123995780944824,
-        "time_taken": 22.13362233299995
+        "accuracy": 0.6276999711990356,
+        "time_taken": 15.224367368999992
       }
     },
     "100": {
       "0": {
-        "accuracy": 0.7077999114990234,
-        "time_taken": 10.597413770000003
+        "accuracy": 0.704299807548523,
+        "time_taken": 5.593184323999992
       },
       "1": {
-        "accuracy": 0.7308999300003052,
-        "time_taken": 8.02459470000008
+        "accuracy": 0.7321001887321472,
+        "time_taken": 5.385782182999947
       },
       "2": {
-        "accuracy": 0.6015002131462097,
-        "time_taken": 3.7315618499999346
+        "accuracy": 0.7243001461029053,
+        "time_taken": 9.138968591000094
       },
       "3": {
-        "accuracy": 0.7232002019882202,
-        "time_taken": 6.287373247999994
+        "accuracy": 0.7279003262519836,
+        "time_taken": 3.660555053000053
       },
       "4": {
-        "accuracy": 0.6919002532958984,
-        "time_taken": 8.955407454000124
+        "accuracy": 0.6853998899459839,
+        "time_taken": 6.721555722999938
       }
     },
     "500": {
       "0": {
-        "accuracy": 0.8521634936332703,
-        "time_taken": 5.262510822999957
+        "accuracy": 0.849459171295166,
+        "time_taken": 4.1964316430000395
       },
       "1": {
-        "accuracy": 0.8403445482254028,
-        "time_taken": 3.346613910999963
+        "accuracy": 0.8390424847602844,
+        "time_taken": 2.6543577679999544
       },
       "2": {
-        "accuracy": 0.8541666865348816,
-        "time_taken": 3.4243159590000687
+        "accuracy": 0.8586738705635071,
+        "time_taken": 3.7907679610000287
       },
       "3": {
-        "accuracy": 0.8365384936332703,
-        "time_taken": 3.701081079000005
+        "accuracy": 0.8433493971824646,
+        "time_taken": 3.211584036999966
       },
       "4": {
-        "accuracy": 0.8471554517745972,
-        "time_taken": 3.271475813000052
+        "accuracy": 0.8508613705635071,
+        "time_taken": 4.339650910000046
       }
     },
     "1000": {
       "0": {
-        "accuracy": 0.8824118971824646,
-        "time_taken": 4.746199452000042
+        "accuracy": 0.8806089758872986,
+        "time_taken": 3.810672871999998
       },
       "1": {
-        "accuracy": 0.8761017918586731,
-        "time_taken": 3.0665597200001002
+        "accuracy": 0.8755007982254028,
+        "time_taken": 2.6573232700000062
       },
       "2": {
-        "accuracy": 0.8830128312110901,
-        "time_taken": 2.602093793999984
+        "accuracy": 0.8828125,
+        "time_taken": 2.6886053809998884
       },
       "3": {
-        "accuracy": 0.8777043223381042,
-        "time_taken": 4.245425575000127
+        "accuracy": 0.8731971383094788,
+        "time_taken": 2.8564365010001893
       },
       "4": {
-        "accuracy": 0.8834134936332703,
-        "time_taken": 2.963790925000012
+        "accuracy": 0.8818109035491943,
+        "time_taken": 3.027272230000108
       }
     },
     "5000": {
       "0": {
-        "accuracy": 0.9229767918586731,
-        "time_taken": 5.274251015000004
+        "accuracy": 0.9258814454078674,
+        "time_taken": 4.428039572999978
       },
       "1": {
-        "accuracy": 0.926682710647583,
-        "time_taken": 5.173077755999998
+        "accuracy": 0.9238781929016113,
+        "time_taken": 3.322857870000007
       },
       "2": {
-        "accuracy": 0.9291867017745972,
-        "time_taken": 4.965678029000173
+        "accuracy": 0.9277844429016113,
+        "time_taken": 3.3129011699998046
       },
       "3": {
-        "accuracy": 0.9264823794364929,
-        "time_taken": 4.216483555999957
+        "accuracy": 0.9291867017745972,
+        "time_taken": 4.800562907000085
       },
       "4": {
-        "accuracy": 0.9259815812110901,
-        "time_taken": 4.042372738000267
+        "accuracy": 0.9294871687889099,
+        "time_taken": 3.492984100000058
       }
     }
   },
   "RPCholesky": {
     "25": {
       "0": {
-        "accuracy": 0.5006002187728882,
-        "time_taken": 41.42257340599997
+        "accuracy": 0.47358962893486023,
+        "time_taken": 19.895207870000036
       },
       "1": {
-        "accuracy": 0.556322455406189,
-        "time_taken": 25.34968879200005
+        "accuracy": 0.5369148850440979,
+        "time_taken": 14.838139795000075
       },
       "2": {
-        "accuracy": 0.5873348712921143,
-        "time_taken": 37.2711402450002
+        "accuracy": 0.5467190146446228,
+        "time_taken": 17.321421823000037
       },
       "3": {
-        "accuracy": 0.45908376574516296,
-        "time_taken": 41.109968239999944
+        "accuracy": 0.4283711016178131,
+        "time_taken": 20.84139633099994
       },
       "4": {
-        "accuracy": 0.47118860483169556,
-        "time_taken": 41.38185949600029
+        "accuracy": 0.4912963807582855,
+        "time_taken": 17.36893893699994
       }
     },
     "50": {
       "0": {
-        "accuracy": 0.5983994603157043,
-        "time_taken": 16.419000583999946
+        "accuracy": 0.625399649143219,
+        "time_taken": 13.748352533000002
       },
       "1": {
-        "accuracy": 0.6796995997428894,
-        "time_taken": 15.655551776000038
+        "accuracy": 0.6291998624801636,
+        "time_taken": 8.75706090899996
       },
       "2": {
-        "accuracy": 0.5455998182296753,
-        "time_taken": 23.019029731000046
+        "accuracy": 0.5662998557090759,
+        "time_taken": 6.946477493999964
       },
       "3": {
-        "accuracy": 0.6141995191574097,
-        "time_taken": 20.926626421000037
+        "accuracy": 0.6297994256019592,
+        "time_taken": 11.587993342000118
       },
       "4": {
-        "accuracy": 0.5852997303009033,
-        "time_taken": 19.28357562400015
+        "accuracy": 0.5483998656272888,
+        "time_taken": 14.694229403999998
       }
     },
     "100": {
       "0": {
-        "accuracy": 0.7150002121925354,
-        "time_taken": 10.279756443999986
+        "accuracy": 0.725000262260437,
+        "time_taken": 8.696777773000008
       },
       "1": {
-        "accuracy": 0.695099949836731,
-        "time_taken": 9.367277195000042
+        "accuracy": 0.7001000046730042,
+        "time_taken": 5.543955067999946
       },
       "2": {
-        "accuracy": 0.6307001709938049,
-        "time_taken": 6.254802073000064
+        "accuracy": 0.6311002373695374,
+        "time_taken": 4.008666183000059
       },
       "3": {
-        "accuracy": 0.671200156211853,
-        "time_taken": 9.271331661999966
+        "accuracy": 0.6850997805595398,
+        "time_taken": 7.576385713000036
       },
       "4": {
-        "accuracy": 0.6927000284194946,
-        "time_taken": 10.04333458400015
+        "accuracy": 0.646899938583374,
+        "time_taken": 9.464091529000143
       }
     },
     "500": {
       "0": {
-        "accuracy": 0.8525640964508057,
-        "time_taken": 6.090376355999979
+        "accuracy": 0.8492588400840759,
+        "time_taken": 5.850811007999994
       },
       "1": {
-        "accuracy": 0.8218148946762085,
-        "time_taken": 3.980846933000066
+        "accuracy": 0.8279246687889099,
+        "time_taken": 3.400919842999997
       },
       "2": {
-        "accuracy": 0.8170072436332703,
-        "time_taken": 5.630132798999966
+        "accuracy": 0.8026843070983887,
+        "time_taken": 4.008895577999965
       },
       "3": {
-        "accuracy": 0.8092948794364929,
-        "time_taken": 4.590955021999889
+        "accuracy": 0.8110977411270142,
+        "time_taken": 4.032388095999977
       },
       "4": {
-        "accuracy": 0.840745210647583,
-        "time_taken": 4.741569309999704
+        "accuracy": 0.8451522588729858,
+        "time_taken": 3.68542980899997
       }
     },
     "1000": {
       "0": {
-        "accuracy": 0.8627804517745972,
-        "time_taken": 5.964579186000037
+        "accuracy": 0.8804086446762085,
+        "time_taken": 5.837933116000045
       },
       "1": {
-        "accuracy": 0.8595753312110901,
-        "time_taken": 4.2806823729999905
+        "accuracy": 0.8630809187889099,
+        "time_taken": 4.17050370100003
       },
       "2": {
-        "accuracy": 0.8712940812110901,
-        "time_taken": 5.8189128090000395
+        "accuracy": 0.8666867017745972,
+        "time_taken": 4.835822229000087
       },
       "3": {
-        "accuracy": 0.8412460088729858,
-        "time_taken": 4.6377100379997955
+        "accuracy": 0.8480569124221802,
+        "time_taken": 4.71791837700016
       },
       "4": {
-        "accuracy": 0.8733974695205688,
-        "time_taken": 4.602818280000065
+        "accuracy": 0.8788061141967773,
+        "time_taken": 4.380343813000081
       }
     },
     "5000": {
       "0": {
-        "accuracy": 0.9246795177459717,
-        "time_taken": 30.759691682999915
+        "accuracy": 0.9239783883094788,
+        "time_taken": 29.51819806499998
       },
       "1": {
-        "accuracy": 0.9284855723381042,
-        "time_taken": 28.251680083999872
+        "accuracy": 0.9287860989570618,
+        "time_taken": 27.981147562000046
       },
       "2": {
-        "accuracy": 0.9284855723381042,
-        "time_taken": 28.629509002000077
+        "accuracy": 0.9247796535491943,
+        "time_taken": 27.539323438999872
       },
       "3": {
-        "accuracy": 0.9278846383094788,
-        "time_taken": 29.119471999000098
+        "accuracy": 0.9285857677459717,
+        "time_taken": 27.830323820999865
       },
       "4": {
-        "accuracy": 0.9231771230697632,
-        "time_taken": 28.0593601999999
+        "accuracy": 0.9238781929016113,
+        "time_taken": 28.524467420000065
       }
     }
   },
   "KernelHerding": {
     "25": {
       "0": {
-        "accuracy": 0.5423170924186707,
-        "time_taken": 52.57508162400006
+        "accuracy": 0.5501202940940857,
+        "time_taken": 29.94118039799997
       },
       "1": {
-        "accuracy": 0.5205084085464478,
-        "time_taken": 39.45232254999996
+        "accuracy": 0.5024006366729736,
+        "time_taken": 19.38333930700003
       },
       "2": {
-        "accuracy": 0.46438610553741455,
-        "time_taken": 35.902196312000115
+        "accuracy": 0.5388156771659851,
+        "time_taken": 16.329524966000008
       },
       "3": {
-        "accuracy": 0.5081029534339905,
-        "time_taken": 44.60431595299997
+        "accuracy": 0.5127052068710327,
+        "time_taken": 16.15010204400005
       },
       "4": {
-        "accuracy": 0.5087032914161682,
-        "time_taken": 41.437250334000055
+        "accuracy": 0.539915919303894,
+        "time_taken": 17.088837638000086
       }
     },
     "50": {
       "0": {
-        "accuracy": 0.6092994213104248,
-        "time_taken": 20.360353013999998
+        "accuracy": 0.6299993395805359,
+        "time_taken": 17.985974345999978
       },
       "1": {
-        "accuracy": 0.6002998352050781,
-        "time_taken": 14.649381372000107
+        "accuracy": 0.6076997518539429,
+        "time_taken": 10.080095013000005
       },
       "2": {
-        "accuracy": 0.5516994595527649,
-        "time_taken": 9.946566445000144
+        "accuracy": 0.6290996670722961,
+        "time_taken": 10.87603085700016
       },
       "3": {
-        "accuracy": 0.5886996984481812,
-        "time_taken": 17.334741768999947
+        "accuracy": 0.6168995499610901,
+        "time_taken": 12.538436354000169
       },
       "4": {
-        "accuracy": 0.6138997673988342,
-        "time_taken": 17.77710858399996
+        "accuracy": 0.5985994935035706,
+        "time_taken": 9.817184417000135
       }
     },
     "100": {
       "0": {
-        "accuracy": 0.7017998099327087,
-        "time_taken": 9.876689930999987
+        "accuracy": 0.7302002310752869,
+        "time_taken": 13.26493101899996
       },
       "1": {
-        "accuracy": 0.6914001107215881,
-        "time_taken": 5.265936994000185
+        "accuracy": 0.7350001931190491,
+        "time_taken": 7.117937374000007
       },
       "2": {
-        "accuracy": 0.7000998258590698,
-        "time_taken": 5.393372559999989
+        "accuracy": 0.723000168800354,
+        "time_taken": 6.991755536000028
       },
       "3": {
-        "accuracy": 0.6898000836372375,
-        "time_taken": 6.27198206900016
+        "accuracy": 0.7188998460769653,
+        "time_taken": 6.8836636179999005
       },
       "4": {
-        "accuracy": 0.698900043964386,
-        "time_taken": 4.835163652000119
+        "accuracy": 0.716400146484375,
+        "time_taken": 5.2314696280000135
       }
     },
     "500": {
       "0": {
-        "accuracy": 0.8225160241127014,
-        "time_taken": 7.761980211000036
+        "accuracy": 0.8127003312110901,
+        "time_taken": 8.29185179000001
       },
       "1": {
-        "accuracy": 0.8360376954078674,
-        "time_taken": 4.213319348000141
+        "accuracy": 0.8148037195205688,
+        "time_taken": 3.6316947710000704
       },
       "2": {
-        "accuracy": 0.8356370329856873,
-        "time_taken": 3.8930105310000727
+        "accuracy": 0.8196113705635071,
+        "time_taken": 3.252079743999957
       },
       "3": {
-        "accuracy": 0.810396671295166,
-        "time_taken": 3.072564359000353
+        "accuracy": 0.8093950152397156,
+        "time_taken": 3.455513597000163
       },
       "4": {
-        "accuracy": 0.8274238705635071,
-        "time_taken": 4.178343173999565
+        "accuracy": 0.8156049847602844,
+        "time_taken": 3.3011748870001156
       }
     },
     "1000": {
       "0": {
-        "accuracy": 0.8530648946762085,
-        "time_taken": 6.340398364999942
+        "accuracy": 0.8576722741127014,
+        "time_taken": 8.381767883000009
       },
       "1": {
-        "accuracy": 0.8548678159713745,
-        "time_taken": 3.055124343999978
+        "accuracy": 0.8568710088729858,
+        "time_taken": 3.56256258399992
       },
       "2": {
-        "accuracy": 0.8617788553237915,
-        "time_taken": 3.762878829000101
+        "accuracy": 0.8465545177459717,
+        "time_taken": 3.135869393999883
       },
       "3": {
-        "accuracy": 0.8614783883094788,
-        "time_taken": 3.4879642840000997
+        "accuracy": 0.860276460647583,
+        "time_taken": 3.4447939570000017
       },
       "4": {
-        "accuracy": 0.8601762652397156,
-        "time_taken": 3.4571287429998847
+        "accuracy": 0.8583734035491943,
+        "time_taken": 3.61459772000012
       }
     },
     "5000": {
       "0": {
-        "accuracy": 0.9297876954078674,
-        "time_taken": 7.996410788999924
+        "accuracy": 0.9268830418586731,
+        "time_taken": 9.059691469000086
       },
       "1": {
-        "accuracy": 0.9332932829856873,
-        "time_taken": 6.031857453999919
+        "accuracy": 0.9315905570983887,
+        "time_taken": 5.148555465000072
       },
       "2": {
-        "accuracy": 0.9241787195205688,
-        "time_taken": 4.493304755000054
+        "accuracy": 0.930588960647583,
+        "time_taken": 5.6641410270001415
       },
       "3": {
-        "accuracy": 0.9307892918586731,
-        "time_taken": 6.0171785980001005
+        "accuracy": 0.9274839758872986,
+        "time_taken": 5.935124415000018
       },
       "4": {
-        "accuracy": 0.9332932829856873,
-        "time_taken": 6.450509147000048
+        "accuracy": 0.9286859035491943,
+        "time_taken": 4.589270231
       }
     }
   },
   "SteinThinning": {
     "25": {
       "0": {
-        "accuracy": 0.38285332918167114,
-        "time_taken": 61.79404670500003
+        "accuracy": 0.36124444007873535,
+        "time_taken": 32.886642288000075
       },
       "1": {
-        "accuracy": 0.37535014748573303,
-        "time_taken": 55.07096734299989
+        "accuracy": 0.3338334560394287,
+        "time_taken": 41.821553507999965
       },
       "2": {
-        "accuracy": 0.38295334577560425,
-        "time_taken": 59.987997941
+        "accuracy": 0.31372541189193726,
+        "time_taken": 34.356144773000096
       },
       "3": {
-        "accuracy": 0.3390357494354248,
-        "time_taken": 40.479311096999936
+        "accuracy": 0.27040842175483704,
+        "time_taken": 33.386861657000054
       },
       "4": {
-        "accuracy": 0.36194491386413574,
-        "time_taken": 39.54609884499996
+        "accuracy": 0.35444176197052,
+        "time_taken": 36.694613785
       }
     },
     "50": {
       "0": {
-        "accuracy": 0.4300001561641693,
-        "time_taken": 33.57654272299999
+        "accuracy": 0.46370017528533936,
+        "time_taken": 37.51419123599999
       },
       "1": {
-        "accuracy": 0.43349993228912354,
-        "time_taken": 31.334013471999924
+        "accuracy": 0.43250009417533875,
+        "time_taken": 33.460962724999945
       },
       "2": {
-        "accuracy": 0.4227999746799469,
-        "time_taken": 32.19259820599996
+        "accuracy": 0.4118999242782593,
+        "time_taken": 29.25043843399999
       },
       "3": {
-        "accuracy": 0.39500007033348083,
-        "time_taken": 33.17475527899978
+        "accuracy": 0.40100017189979553,
+        "time_taken": 27.707170819999874
       },
       "4": {
-        "accuracy": 0.4095003008842468,
-        "time_taken": 35.54211925799973
+        "accuracy": 0.45050016045570374,
+        "time_taken": 28.452544991999957
       }
     },
     "100": {
       "0": {
-        "accuracy": 0.4675999581813812,
-        "time_taken": 31.202895077999983
+        "accuracy": 0.4580000340938568,
+        "time_taken": 23.420479623999995
       },
       "1": {
-        "accuracy": 0.49570003151893616,
-        "time_taken": 27.87190726899985
+        "accuracy": 0.4966999590396881,
+        "time_taken": 26.137647441000013
       },
       "2": {
-        "accuracy": 0.47530001401901245,
-        "time_taken": 26.891571627000076
+        "accuracy": 0.500999927520752,
+        "time_taken": 24.4042449760002
       },
       "3": {
-        "accuracy": 0.48059985041618347,
-        "time_taken": 24.772397474999707
+        "accuracy": 0.4515998661518097,
+        "time_taken": 22.571522415000118
       },
       "4": {
-        "accuracy": 0.44819989800453186,
-        "time_taken": 24.519068094999966
+        "accuracy": 0.4860999286174774,
+        "time_taken": 27.187762750000047
       }
     },
     "500": {
       "0": {
-        "accuracy": 0.5233373641967773,
-        "time_taken": 32.518364321000035
+        "accuracy": 0.5588942170143127,
+        "time_taken": 20.343176127999982
       },
       "1": {
-        "accuracy": 0.5446714758872986,
-        "time_taken": 21.10401904199989
+        "accuracy": 0.5828325152397156,
+        "time_taken": 19.380917395999973
       },
       "2": {
-        "accuracy": 0.5258413553237915,
-        "time_taken": 21.467372049999994
+        "accuracy": 0.5852363705635071,
+        "time_taken": 19.87485118199993
       },
       "3": {
-        "accuracy": 0.5215344429016113,
-        "time_taken": 21.456703629999993
+        "accuracy": 0.5499799847602844,
+        "time_taken": 22.195402222999974
       },
       "4": {
-        "accuracy": 0.5278445482254028,
-        "time_taken": 22.62931420799987
+        "accuracy": 0.5618990659713745,
+        "time_taken": 19.422155341999996
       }
     },
     "1000": {
       "0": {
-        "accuracy": 0.5904447436332703,
-        "time_taken": 27.6814611100001
+        "accuracy": 0.5892428159713745,
+        "time_taken": 21.69759744099997
       },
       "1": {
-        "accuracy": 0.5819311141967773,
-        "time_taken": 24.345591108000008
+        "accuracy": 0.5992588400840759,
+        "time_taken": 22.000713124999947
       },
       "2": {
-        "accuracy": 0.5763221383094788,
-        "time_taken": 22.27670924300014
+        "accuracy": 0.5953525900840759,
+        "time_taken": 23.080259337999905
       },
       "3": {
-        "accuracy": 0.5704126954078674,
-        "time_taken": 25.645224296999913
+        "accuracy": 0.5842347741127014,
+        "time_taken": 20.16238621299999
       },
       "4": {
-        "accuracy": 0.5785256624221802,
-        "time_taken": 22.827015231999667
+        "accuracy": 0.5821314454078674,
+        "time_taken": 20.503732982999963
       }
     },
     "5000": {
       "0": {
-        "accuracy": 0.6521434187889099,
-        "time_taken": 38.882569371000045
+        "accuracy": 0.6593549847602844,
+        "time_taken": 29.339991162000047
       },
       "1": {
-        "accuracy": 0.6533453464508057,
-        "time_taken": 34.77338421800005
+        "accuracy": 0.6642628312110901,
+        "time_taken": 29.7060228900001
       },
       "2": {
-        "accuracy": 0.6462339758872986,
-        "time_taken": 37.378769906000116
+        "accuracy": 0.6545472741127014,
+        "time_taken": 30.334544582000035
       },
       "3": {
-        "accuracy": 0.6377203464508057,
-        "time_taken": 34.397472245000245
+        "accuracy": 0.6499398946762085,
+        "time_taken": 29.475541733
       },
       "4": {
-        "accuracy": 0.6459335088729858,
-        "time_taken": 38.09663688599994
+        "accuracy": 0.6581530570983887,
+        "time_taken": 30.22798960299997
       }
     }
   }
-}
+}
\ No newline at end of file

From 874467ad780c36e3576cb482d90c2fe7ef97ed20 Mon Sep 17 00:00:00 2001
From: qh681248 <181246904+qh681248@users.noreply.github.com>
Date: Thu, 19 Dec 2024 15:23:26 +0000
Subject: [PATCH 8/8] feat: used path lib consistently #893

---
 benchmark/mnist_benchmark_results.json |  2 +-
 benchmark/pounce_benchmark.py          | 11 +++++------
 2 files changed, 6 insertions(+), 7 deletions(-)

diff --git a/benchmark/mnist_benchmark_results.json b/benchmark/mnist_benchmark_results.json
index 456c1dd0..63d18543 100644
--- a/benchmark/mnist_benchmark_results.json
+++ b/benchmark/mnist_benchmark_results.json
@@ -535,4 +535,4 @@
       }
     }
   }
-}
\ No newline at end of file
+}
diff --git a/benchmark/pounce_benchmark.py b/benchmark/pounce_benchmark.py
index 1d3d8c77..d5ee7801 100644
--- a/benchmark/pounce_benchmark.py
+++ b/benchmark/pounce_benchmark.py
@@ -21,11 +21,11 @@
 4. Print the time taken to generate each coreset.
 """
 
-import os
 import time
 from pathlib import Path
 
 import imageio
+import jax.numpy as jnp
 import numpy as np
 import umap
 from jax import random
@@ -46,18 +46,17 @@ def benchmark_coreset_algorithms(
     :param out_dir: Directory to save the output GIFs for each coreset algorithm.
     :param coreset_size: The size of the coreset.
     """
-    base_dir = os.path.dirname(os.path.abspath(__file__))
+    base_dir = Path(__file__).resolve().parent
 
     # Ensure paths are absolute and output directory exists
-    in_path = Path(os.path.join(base_dir, in_path)).resolve()
-    out_dir = Path(os.path.join(base_dir, out_dir)).resolve()
+    in_path = (base_dir / in_path).resolve()
+    out_dir = (base_dir / out_dir).resolve()
     out_dir.mkdir(parents=True, exist_ok=True)
 
     # Load and preprocess video frames
     _, *image_data = imageio.v2.mimread(in_path)
     raw_data = np.asarray(image_data)
     reshaped_data = raw_data.reshape(raw_data.shape[0], -1)
-    print(type(reshaped_data))
 
     umap_model = umap.UMAP(densmap=True, n_components=25)
     umap_data = umap_model.fit_transform(reshaped_data)
@@ -71,7 +70,7 @@ def benchmark_coreset_algorithms(
     for get_solver in solvers:
         solver = get_solver(coreset_size)
         solver_name = get_solver_name(solver)
-        data = Data(umap_data)
+        data = Data(jnp.array(umap_data))
 
         start_time = time.perf_counter()
         coreset, _ = solver.reduce(data)