[pre-commit.ci] pre-commit autoupdate

.pre-commit-config.yaml

@@ -1,6 +1,6 @@
 repos:
 - repo: https://github.com/astral-sh/ruff-pre-commit
-  rev: v0.8.6
+  rev: v0.9.1
   hooks:
   - id: ruff
     args:

happypose/pose_estimators/cosypose/cosypose/config.py

@@ -21,9 +21,9 @@
 LOCAL_DATA_DIR = Path(
     os.environ.get("HAPPYPOSE_DATA_DIR", Path(PROJECT_DIR) / "local_data"),
 )
-assert (
-    LOCAL_DATA_DIR.exists()
-), "Did you forget to set env variable 'HAPPYPOSE_DATA_DIR'?"
+assert LOCAL_DATA_DIR.exists(), (
+    "Did you forget to set env variable 'HAPPYPOSE_DATA_DIR'?"
+)
 TEST_DATA_DIR = LOCAL_DATA_DIR
 DASK_LOGS_DIR = LOCAL_DATA_DIR / "dasklogs"
 SYNT_DS_DIR = LOCAL_DATA_DIR / "synt_datasets"

happypose/pose_estimators/cosypose/cosypose/evaluation/pred_runner/bop_predictions.py

@@ -103,7 +103,7 @@ def get_predictions(
             depth = None
             if self.load_depth:
                 depth = data["depth"].cuda().float()
-            logger.debug(f"{'-'*80}")
+            logger.debug(f"{'-' * 80}")
             logger.debug(f"Predictions on {data['im_infos']}")
 
             def get_preds():
@@ -155,7 +155,7 @@ def get_preds():
                     f"Full predictions: {n_dets} detections + pose estimation "
                     f"in {duration:.3f} s",
                 )
-                logger.debug(f"{'-'*80}")
+                logger.debug(f"{'-' * 80}")
                 return this_batch_detections, all_preds, duration
 
             # Run once without measuring timing

happypose/pose_estimators/cosypose/cosypose/evaluation/pred_runner/multiview_predictions.py

@@ -118,7 +118,7 @@ def get_predictions(
             view_ids = np.unique(gt_detections.infos["view_id"])
             group_id = np.unique(gt_detections.infos["group_id"])
             n_gt_dets = len(gt_detections)
-            logger.debug(f"{'-'*80}")
+            logger.debug(f"{'-' * 80}")
             logger.debug(f"Scene: {scene_id}")
             logger.debug(f"Views: {view_ids}")
             logger.debug(f"Group: {group_id}")
@@ -163,7 +163,7 @@ def get_predictions(
                         candidates,
                         cameras,
                     )
-            logger.debug(f"{'-'*80}")
+            logger.debug(f"{'-' * 80}")
 
             for k, v in sv_preds.items():
                 predictions[k].append(v.cpu())

happypose/pose_estimators/cosypose/cosypose/evaluation/runner_utils.py

@@ -49,14 +49,14 @@ def format_results(predictions, eval_metrics, eval_dfs, print_metrics=True):
     df = defaultdict(list)
     summary_txt = ""
     for k, v in eval_metrics.items():
-        summary_txt += f"\n{k}\n{'-'*80}\n"
+        summary_txt += f"\n{k}\n{'-' * 80}\n"
         for k_, v_ in v.items():
             summary[f"{k}/{k_}"] = v_
             df["method"].append(k)
             df["metric"].append(k_)
             df["value"].append(v_)
             summary_txt += f"{k}/{k_}: {v_}\n"
-        summary_txt += f"{'-'*80}"
+        summary_txt += f"{'-' * 80}"
     if print_metrics:
         logger.info(summary_txt)
 

happypose/pose_estimators/cosypose/cosypose/integrated/pose_estimator.py

@@ -161,7 +161,7 @@ def run_inference_pipeline(
 
         if coarse_estimates is None:
             assert detections is not None or run_detector, (
-                "You must " "either pass in `detections` or set run_detector=True"
+                "You must either pass in `detections` or set run_detector=True"
             )
             if detections is None and run_detector:
                 start_time = time.time()

happypose/pose_estimators/cosypose/cosypose/models/pose.py

@@ -161,7 +161,7 @@ def forward(self, images, K, labels, TCO, n_iterations=1):
 
             TCO_output = self.update_pose(TCO_input, K_crop, model_outputs["pose"])
 
-            outputs[f"iteration={n+1}"] = {
+            outputs[f"iteration={n + 1}"] = {
                 "TCO_input": TCO_input,
                 "TCO_output": TCO_output,
                 "K_crop": K_crop,
@@ -170,7 +170,7 @@ def forward(self, images, K, labels, TCO, n_iterations=1):
                 "boxes_crop": boxes_crop,
             }
 
-            outputs[f"iteration={n+1}"] = PosePredictorOutputCosypose(
+            outputs[f"iteration={n + 1}"] = PosePredictorOutputCosypose(
                 renders=renders.rgbs,
                 images_crop=images_crop,
                 TCO_input=TCO_input,
@@ -186,13 +186,13 @@ def forward(self, images, K, labels, TCO, n_iterations=1):
             TCO_input = TCO_output
 
             if self.debug:
-                self.tmp_debug.update(outputs[f"iteration={n+1}"])
+                self.tmp_debug.update(outputs[f"iteration={n + 1}"])
                 self.tmp_debug.update(
                     images=images,
                     images_crop=images_crop,
                     renders=renders.rgbs,
                 )
-                path = DEBUG_DATA_DIR / f"debug_iter={n+1}.pth.tar"
+                path = DEBUG_DATA_DIR / f"debug_iter={n + 1}.pth.tar"
                 logger.info(f"Wrote debug data: {path}")
                 torch.save(self.tmp_debug, path)
 

happypose/pose_estimators/cosypose/cosypose/scripts/run_bop20_eval_multi.py

@@ -55,14 +55,14 @@ def main():
         proc.join()
 
     for _ in range(10):
-        print(f"{'-'*80}")
+        print(f"{'-' * 80}")
 
     for ds_name, (cfg, _) in processes.items():
         results_dir = (
             LOCAL_DATA_DIR / "bop_eval_outputs" / cfg.csv_path.with_suffix("").name
         )
         scores_path = results_dir / "scores_bop19.json"
-        print(f"{'-'*80}")
+        print(f"{'-' * 80}")
         print(f"{ds_name}: {scores_path}")
         print(scores_path.read_text())
 

happypose/pose_estimators/cosypose/cosypose/scripts/run_bop_inference.py

@@ -117,10 +117,10 @@ def load_pose_models(coarse_run_id, refiner_run_id=None, n_workers=8):
 
 
 def run_inference(args):
-    logger.info(f"{'-'*80}")
+    logger.info(f"{'-' * 80}")
     for k, v in args.__dict__.items():
         logger.info(f"{k}: {v}")
-    logger.info(f"{'-'*80}")
+    logger.info(f"{'-' * 80}")
 
     scene_ds = make_scene_dataset(args.ds_name, n_frames=args.n_frames)
 

happypose/pose_estimators/cosypose/cosypose/scripts/run_cosypose_eval.py

@@ -527,8 +527,7 @@ def main():
             f"CLASS/AUC/objects/mean"
         )
         k_4 = (
-            f"{det_key}/ba_output+all_cand/ADD-S_ntop=1_matching="
-            f"CLASS/AUC/objects/mean"
+            f"{det_key}/ba_output+all_cand/ADD-S_ntop=1_matching=CLASS/AUC/objects/mean"
         )
         metrics_to_print.update(
             {
@@ -595,9 +594,9 @@ def main():
         for k, v in metrics_to_print.items():
             if k in full_summary:
                 summary_txt += f"\n{v}: {full_summary[k]}"
-        logger.info(f"{'-'*80}")
+        logger.info(f"{'-' * 80}")
         logger.info(summary_txt)
-        logger.info(f"{'-'*80}")
+        logger.info(f"{'-' * 80}")
 
         torch.save(results, save_dir / "results.pth.tar")
         (save_dir / "summary.txt").write_text(summary_txt)

happypose/pose_estimators/cosypose/cosypose/scripts/run_custom_scenario.py

@@ -161,19 +161,19 @@ def main():
         if "cosypose" in logger.name:
             logger.setLevel(logging.DEBUG)
 
-    logger.info(f"{'-'*80}")
+    logger.info(f"{'-' * 80}")
     for k, v in args.__dict__.items():
         logger.info(f"{k}: {v}")
-    logger.info(f"{'-'*80}")
+    logger.info(f"{'-' * 80}")
 
     scenario_dir = LOCAL_DATA_DIR / "custom_scenarios" / args.scenario
 
     candidates = read_csv_candidates(scenario_dir / "candidates.csv").float().cuda()
     candidates.infos["group_id"] = 0
     scene_ids = np.unique(candidates.infos["scene_id"])
-    assert (
-        len(scene_ids) == 1
-    ), "Please only provide 6D pose estimations that correspond to the same scene."
+    assert len(scene_ids) == 1, (
+        "Please only provide 6D pose estimations that correspond to the same scene."
+    )
     scene_id = scene_ids.item()
     view_ids = np.unique(candidates.infos["view_id"])
     n_views = len(view_ids)

happypose/pose_estimators/cosypose/cosypose/scripts/run_detection_eval.py

@@ -93,10 +93,10 @@ def get_meters(scene_ds):
 
 
 def run_detection_eval(args, detector=None):
-    logger.info(f"{'-'*80}")
+    logger.info(f"{'-' * 80}")
     for k, v in args.__dict__.items():
         logger.info(f"{k}: {v}")
-    logger.info(f"{'-'*80}")
+    logger.info(f"{'-' * 80}")
 
     scene_ds = make_scene_dataset(args.ds_name, n_frames=args.n_frames)
 

happypose/pose_estimators/cosypose/cosypose/scripts/run_inference_on_example.py

@@ -74,9 +74,9 @@ def main():
     data_dir = os.getenv("HAPPYPOSE_DATA_DIR")
     assert data_dir, "Set HAPPYPOSE_DATA_DIR env variable"
     example_dir = Path(data_dir) / "examples" / args.example_name
-    assert (
-        example_dir.exists()
-    ), "Example {args.example_name} not available, follow download instructions"
+    assert example_dir.exists(), (
+        "Example {args.example_name} not available, follow download instructions"
+    )
     # dataset_to_use = args.dataset  # hope/tless/ycbv
 
     # Load data

happypose/pose_estimators/cosypose/cosypose/simulator/caching.py

@@ -17,7 +17,7 @@ def __init__(self, urdf_ds, client_id):
     def _load_body(self, label):
         ds_idx = np.where(self.urdf_ds.index["label"] == label)[0]
         if len(ds_idx) == 0:
-            raise ValueError(f'Label {label} not in {self.urdf_ds.index["label"]}')
+            raise ValueError(f"Label {label} not in {self.urdf_ds.index['label']}")
         object_infos = self.urdf_ds[ds_idx.item()].to_dict()
         body = Body.load(
             object_infos["urdf_path"],

happypose/pose_estimators/cosypose/cosypose/training/pose_forward_loss.py

@@ -62,7 +62,7 @@ def h_pose(model, mesh_db, data, meters, cfg, n_iterations=1, input_generator="f
 
     losses_TCO_iter = []
     for n in range(n_iterations):
-        iter_outputs = outputs[f"iteration={n+1}"]
+        iter_outputs = outputs[f"iteration={n + 1}"]
         K_crop = iter_outputs.K_crop
         TCO_input = iter_outputs.TCO_input
         TCO_pred = iter_outputs.TCO_output
@@ -90,7 +90,7 @@ def h_pose(model, mesh_db, data, meters, cfg, n_iterations=1, input_generator="f
                 points,
             )
 
-        meters[f"loss_TCO-iter={n+1}"].add(loss_TCO_iter.mean().item())
+        meters[f"loss_TCO-iter={n + 1}"].add(loss_TCO_iter.mean().item())
         losses_TCO_iter.append(loss_TCO_iter)
 
     loss_TCO = torch.cat(losses_TCO_iter).mean()

happypose/pose_estimators/cosypose/cosypose/training/train_detector.py

@@ -129,10 +129,10 @@ def train_detector(args):
     args = check_update_config(args)
     args.save_dir = EXP_DIR / args.run_id
 
-    logger.info(f"{'-'*80}")
+    logger.info(f"{'-' * 80}")
     for k, v in args.__dict__.items():
         logger.info(f"{k}: {v}")
-    logger.info(f"{'-'*80}")
+    logger.info(f"{'-' * 80}")
     # Initialize distributed
     device = torch.cuda.current_device()
     init_distributed_mode()

happypose/pose_estimators/cosypose/cosypose/training/train_pose.py

@@ -265,10 +265,10 @@ def train_pose(args):
     args.save_dir = EXP_DIR / args.run_id
     args = check_update_config(args)
 
-    logger.info(f"{'-'*80}")
+    logger.info(f"{'-' * 80}")
     for k, v in args.__dict__.items():
         logger.info(f"{k}: {v}")
-    logger.info(f"{'-'*80}")
+    logger.info(f"{'-' * 80}")
 
     # Initialize distributed
     device = torch.cuda.current_device()

happypose/pose_estimators/cosypose/cosypose/utils/tensor_collection.py

@@ -39,7 +39,7 @@ def delete_tensor(self, name):
         del self._tensors[name]
 
     def __repr__(self):
-        s = self.__class__.__name__ + "(" "\n"
+        s = self.__class__.__name__ + "(\n"
         for k, t in self._tensors.items():
             s += f"    {k}: {t.shape} {t.dtype} {t.device},\n"
         s += ")"
@@ -131,10 +131,10 @@ def clone(self):
         return PandasTensorCollection(self.infos.copy(), **tensors)
 
     def __repr__(self):
-        s = self.__class__.__name__ + "(" "\n"
+        s = self.__class__.__name__ + "(\n"
         for k, t in self._tensors.items():
             s += f"    {k}: {t.shape} {t.dtype} {t.device},\n"
-        s += f"{'-'*40}\n"
+        s += f"{'-' * 40}\n"
         s += "    infos:\n" + self.infos.__repr__() + "\n"
         s += ")"
         return s

happypose/pose_estimators/megapose/config.py

@@ -31,9 +31,9 @@
 LOCAL_DATA_DIR = Path(
     os.environ.get("HAPPYPOSE_DATA_DIR", Path(PROJECT_DIR) / "local_data"),
 )
-assert (
-    LOCAL_DATA_DIR.exists()
-), "Did you forget to set env variable 'HAPPYPOSE_DATA_DIR'?"
+assert LOCAL_DATA_DIR.exists(), (
+    "Did you forget to set env variable 'HAPPYPOSE_DATA_DIR'?"
+)
 BOP_DS_DIR = LOCAL_DATA_DIR / "bop_datasets"
 NB_DATA_DIR = LOCAL_DATA_DIR / "notebook_data"
 SHAPENET_DIR = LOCAL_DATA_DIR / "shapenetcorev2"

happypose/pose_estimators/megapose/evaluation/runner_utils.py

@@ -63,14 +63,14 @@ def format_results(predictions, eval_metrics, eval_dfs, print_metrics=True):
     df = defaultdict(list)
     summary_txt = ""
     for k, v in eval_metrics.items():
-        summary_txt += f"\n{k}\n{'-'*80}\n"
+        summary_txt += f"\n{k}\n{'-' * 80}\n"
         for k_, v_ in v.items():
             summary[f"{k}/{k_}"] = v_
             df["method"].append(k)
             df["metric"].append(k_)
             df["value"].append(v_)
             summary_txt += f"{k}/{k_}: {v_}\n"
-        summary_txt += f"{'-'*80}"
+        summary_txt += f"{'-' * 80}"
     if print_metrics:
         logger.info(summary_txt)
 

happypose/pose_estimators/megapose/inference/pose_estimator.py

@@ -558,7 +558,7 @@ def run_inference_pipeline(
 
         if coarse_estimates is None:
             assert detections is not None or run_detector, (
-                "You must " "either pass in `detections` or set run_detector=True"
+                "You must either pass in `detections` or set run_detector=True"
             )
             if detections is None and run_detector:
                 start_time = time.time()
@@ -576,9 +576,9 @@ def run_inference_pipeline(
                     labels_to_keep,
                 )
 
-            assert (
-                len(detections) > 0
-            ), "TOFIX: currently, dealing with absence of detections is not supported"
+            assert len(detections) > 0, (
+                "TOFIX: currently, dealing with absence of detections is not supported"
+            )
 
             # Ensure that detections has the instance_id column
             detections = add_instance_id(detections)

happypose/pose_estimators/megapose/models/pose_rigid.py

@@ -651,7 +651,7 @@ def forward(
                     device=device,
                 )
 
-            outputs[f"iteration={n+1}"] = PosePredictorOutput(
+            outputs[f"iteration={n + 1}"] = PosePredictorOutput(
                 renders=renders,
                 images_crop=images_crop,
                 TCO_input=TCO_input,
@@ -669,7 +669,7 @@ def forward(
                 timing_dict=timing_dict,
             )
             if self.debug:
-                self.debug_data.output = outputs[f"iteration={n+1}"]
+                self.debug_data.output = outputs[f"iteration={n + 1}"]
             TCO_input = TCO_output
         return outputs
 
@@ -689,9 +689,9 @@ def forward_coarse_tensor(
                 input + rendered image
 
         """
-        assert (
-            self.predict_rendered_views_logits
-        ), "Method only valid if coarse classification model"
+        assert self.predict_rendered_views_logits, (
+            "Method only valid if coarse classification model"
+        )
         if torch.cuda.is_available():
             timer = CudaTimer(enabled=cuda_timer)
         else:
@@ -738,9 +738,9 @@ def forward_coarse(
 
         """
 
-        assert (
-            self.predict_rendered_views_logits
-        ), "Method only valid if coarse classification model"
+        assert self.predict_rendered_views_logits, (
+            "Method only valid if coarse classification model"
+        )
 
         if not self.input_depth:
             # Remove the depth dimension if it is not used

happypose/pose_estimators/megapose/scripts/make_shapenet_panda3d.py

@@ -102,7 +102,7 @@ def make_new_block():
             face = line_f[3:].split(" ")
             face = [f.split("/") for f in face]
             face_flipped = " ".join(
-                [f"{x[0]}/{x[1]}/{int(x[2])+n_vn_orig}" for x in face],
+                [f"{x[0]}/{x[1]}/{int(x[2]) + n_vn_orig}" for x in face],
             )
             f_flipped.append(f"f  {face_flipped}")
         block["f"] += f_flipped