diff --git a/pyproject.toml b/pyproject.toml
index 872a5d5c5..75ea034fe 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -79,6 +79,7 @@ test = [
     "pre-commit",     # linting
     "pytest",         # unit testing
     "pytest-cov",     # code coverage
+    "pytest-xvfb",    # for headless testing of Qt
 ]
 docs = [
     "hatch",                 # Get version from env
@@ -120,12 +121,12 @@ ignore-words-list = 'nevers'
 [tool.pytest.ini_options]
 minversion = "7.0"
 addopts = [
-    "-sv",
-    "--sw", # stepwise: resume with next test after failure
-    "--pdb", # drop into debugger on failure
+    # "-sv", # verbose output
+    # "--sw", # stepwise: resume with next test after failure
+    # "--pdb", # drop into debugger on failure
     "-p no:warnings",
-    "--no-teardown", # don't teardown the database after tests
-    "--quiet-spy",                  # don't show logging from spyglass
+    # "--no-teardown", # don't teardown the database after tests
+    # "--quiet-spy",                  # don't show logging from spyglass
     "--show-capture=no",
     "--pdbcls=IPython.terminal.debugger:TerminalPdb", # use ipython debugger
     "--cov=spyglass",
@@ -134,6 +135,12 @@ addopts = [
 ]
 testpaths = ["tests"]
 log_level = "INFO"
+env = [
+    "QT_QPA_PLATFORM = offscreen", # QT fails headless without this
+    # "DISPLAY = :0", # QT fails headless without this
+    "TF_ENABLE_ONEDNN_OPTS = 0", # TF disable approx calcs
+    "TF_CPP_MIN_LOG_LEVEL = 2", # Disable TF warnings
+]
 
 [tool.coverage.run]
 source = ["*/src/spyglass/*"]
diff --git a/src/spyglass/decoding/v0/dj_decoder_conversion.py b/src/spyglass/decoding/v0/dj_decoder_conversion.py
index edcb0d637..1cf6d30c4 100644
--- a/src/spyglass/decoding/v0/dj_decoder_conversion.py
+++ b/src/spyglass/decoding/v0/dj_decoder_conversion.py
@@ -26,6 +26,21 @@
         ObservationModel,
     )
 except ImportError as e:
+    (
+        Identity,
+        RandomWalk,
+        RandomWalkDirection1,
+        RandomWalkDirection2,
+        Uniform,
+        DiagonalDiscrete,
+        RandomDiscrete,
+        UniformDiscrete,
+        UserDefinedDiscrete,
+        Environment,
+        UniformInitialConditions,
+        UniformOneEnvironmentInitialConditions,
+        ObservationModel,
+    ) = [None] * 13
     logger.warning(e)
 from track_linearization import make_track_graph
 
diff --git a/src/spyglass/position/v1/dlc_utils.py b/src/spyglass/position/v1/dlc_utils.py
index 7b14efa85..6d27615e4 100644
--- a/src/spyglass/position/v1/dlc_utils.py
+++ b/src/spyglass/position/v1/dlc_utils.py
@@ -385,7 +385,17 @@ def infer_output_dir(key, makedir=True):
     """
     # TODO: add check to make sure interval_list_name refers to a single epoch
     # Or make key include epoch in and of itself instead of interval_list_name
-    nwb_file_name = key["nwb_file_name"].split("_.")[0]
+
+    file_name = key.get("nwb_file_name")
+    dlc_model_name = key.get("dlc_model_name")
+    epoch = key.get("epoch")
+
+    if not all([file_name, dlc_model_name, epoch]):
+        raise ValueError(
+            "Key must contain 'nwb_file_name', 'dlc_model_name', and 'epoch'"
+        )
+
+    nwb_file_name = file_name.split("_.")[0]
     output_dir = pathlib.Path(dlc_output_dir) / pathlib.Path(
         f"{nwb_file_name}/{nwb_file_name}_{key['epoch']:02}"
         f"_model_" + key["dlc_model_name"].replace(" ", "-")
@@ -1021,7 +1031,10 @@ def make_video(
         video.release()
         out.release()
         print("destroying cv2 windows")
-        cv2.destroyAllWindows()
+        try:
+            cv2.destroyAllWindows()
+        except cv2.error:  # if cv is already closed or does not have func
+            pass
         print("finished making video with opencv")
         return
 
diff --git a/src/spyglass/position/v1/position_dlc_pose_estimation.py b/src/spyglass/position/v1/position_dlc_pose_estimation.py
index 42733fd0f..6ae7669bf 100644
--- a/src/spyglass/position/v1/position_dlc_pose_estimation.py
+++ b/src/spyglass/position/v1/position_dlc_pose_estimation.py
@@ -35,7 +35,7 @@ class DLCPoseEstimationSelection(SpyglassMixin, dj.Manual):
     """
 
     @classmethod
-    def get_video_crop(cls, video_path):
+    def get_video_crop(cls, video_path, crop_input=None):
         """
         Queries the user to determine the cropping parameters for a given video
 
@@ -61,9 +61,13 @@ def get_video_crop(cls, video_path):
         ax.set_yticks(np.arange(ylims[0], ylims[-1], -50))
         ax.grid(visible=True, color="white", lw=0.5, alpha=0.5)
         display(fig)
-        crop_input = input(
-            "Please enter the crop parameters for your video in format xmin, xmax, ymin, ymax, or 'none'\n"
-        )
+
+        if crop_input is None:
+            crop_input = input(
+                "Please enter the crop parameters for your video in format "
+                + "xmin, xmax, ymin, ymax, or 'none'\n"
+            )
+
         plt.close()
         if crop_input.lower() == "none":
             return None
diff --git a/src/spyglass/position/v1/position_dlc_selection.py b/src/spyglass/position/v1/position_dlc_selection.py
index 70400dbba..02692ce14 100644
--- a/src/spyglass/position/v1/position_dlc_selection.py
+++ b/src/spyglass/position/v1/position_dlc_selection.py
@@ -304,6 +304,8 @@ class DLCPosVideo(SpyglassMixin, dj.Computed):
     ---
     """
 
+    # TODO: Shoultn't this keep track of the video file it creates?
+
     def make(self, key):
         from tqdm import tqdm as tqdm
 
@@ -432,3 +434,4 @@ def make(self, key):
             crop=crop,
             **params["video_params"],
         )
+        self.insert1(key)
diff --git a/src/spyglass/position/v1/position_dlc_training.py b/src/spyglass/position/v1/position_dlc_training.py
index 6393a8a29..393eb6af9 100644
--- a/src/spyglass/position/v1/position_dlc_training.py
+++ b/src/spyglass/position/v1/position_dlc_training.py
@@ -108,7 +108,7 @@ class DLCModelTrainingSelection(SpyglassMixin, dj.Manual):
     """
 
     def insert1(self, key, **kwargs):
-        training_id = key["training_id"]
+        training_id = key.get("training_id")
         if training_id is None:
             training_id = (
                 dj.U().aggr(self & key, n="max(training_id)").fetch1("n") or 0
diff --git a/tests/README.md b/tests/README.md
index 476dbb4c8..20d4bd2bc 100644
--- a/tests/README.md
+++ b/tests/README.md
@@ -1,5 +1,19 @@
 # PyTests
 
+## Environment
+
+To facilitate headless testing of various Qt-based tools as well as Tensorflow,
+`pyproject.toml` includes some environment variables associated with the
+display. These are...
+
+- `QT_QPA_PLATFORM`: Set to `offscreen` to prevent the need for a display.
+- `TF_ENABLE_ONEDNN_OPTS`: Set to `1` to enable Tensorflow optimizations.
+- `TF_CPP_MIN_LOG_LEVEL`: Set to `2` to suppress Tensorflow warnings.
+
+<!-- - `DISPLAY`: Set to `:0` to prevent the need for a display. -->
+
+## Options
+
 This directory is contains files for testing the code. Simply by running
 `pytest` from the root directory, all tests will be run with default parameters
 specified in `pyproject.toml`. Notable optional parameters include...
@@ -7,7 +21,7 @@ specified in `pyproject.toml`. Notable optional parameters include...
 - Coverage items. The coverage report indicates what percentage of the code was
     included in tests.
 
-    - `--cov=spyglatss`: Which package should be described in the coverage report
+    - `--cov=spyglass`: Which package should be described in the coverage report
     - `--cov-report term-missing`: Include lines of items missing in coverage
 
 - Verbosity.
diff --git a/tests/common/test_behav.py b/tests/common/test_behav.py
index e659cef5c..56ba037b2 100644
--- a/tests/common/test_behav.py
+++ b/tests/common/test_behav.py
@@ -79,11 +79,6 @@ def test_populate_state_script(common, pop_state_script):
     ), "StateScript populate unexpected effect"
 
 
-@pytest.fixture(scope="session")
-def video_keys(common):
-    return common.VideoFile().fetch(as_dict=True)
-
-
 @pytest.mark.usefixtures("skipif_noextras")
 def test_videofile_update_entries(common, video_keys):
     """Test update entries"""
diff --git a/tests/conftest.py b/tests/conftest.py
index 2e2f5633b..05b1a0f72 100644
--- a/tests/conftest.py
+++ b/tests/conftest.py
@@ -388,6 +388,14 @@ def populate_exception():
     yield PopulateException
 
 
+# -------------------------- FIXTURES, COMMON TABLES --------------------------
+
+
+@pytest.fixture(scope="session")
+def video_keys(common):
+    return common.VideoFile().fetch(as_dict=True)
+
+
 # ------------------------- FIXTURES, POSITION TABLES -------------------------
 
 
@@ -428,7 +436,7 @@ def trodes_params(trodes_params_table, teardown):
             },
         },
     }
-    trodes_params_table.get_default()
+    _ = trodes_params_table.get_default()
     trodes_params_table.insert(
         [v for k, v in paramsets.items()], skip_duplicates=True
     )
@@ -778,10 +786,16 @@ def dlc_project_tbl(sgp):
     yield sgp.v1.DLCProject()
 
 
+@pytest.fixture(scope="session")
+def dlc_project_name():
+    yield "pytest_proj"
+
+
 @pytest.fixture(scope="session")
 def insert_project(
     verbose_context,
     teardown,
+    dlc_project_name,
     dlc_project_tbl,
     common,
     bodyparts,
@@ -791,7 +805,7 @@ def insert_project(
     common.LabTeam.insert1({"team_name": team_name}, skip_duplicates=True)
     with verbose_context:
         project_key = dlc_project_tbl.insert_new_project(
-            project_name="pytest_proj",
+            project_name=dlc_project_name,
             bodyparts=bodyparts,
             lab_team=team_name,
             frames_per_video=100,
@@ -858,8 +872,14 @@ def extract_frames(
         )
     vid_name = list(dlc_config["video_sets"].keys())[0].split("/")[-1]
     label_dir = project_dir / "labeled-data" / vid_name.split(".")[0]
+
     yield label_dir
 
+    for file in label_dir.glob("*png"):
+        if file.stem in ["img000", "img001"]:
+            continue
+        file.unlink()
+
 
 @pytest.fixture(scope="session")
 def labeled_vid_dir(extract_frames):
@@ -889,11 +909,18 @@ def add_training_files(dlc_project_tbl, project_key, fix_downloaded):
 
 
 @pytest.fixture(scope="session")
-def training_params_key(verbose_context, sgp, project_key):
-    training_params_name = "pytest"
+def dlc_training_params(sgp):
+    params_tbl = sgp.v1.DLCModelTrainingParams()
+    params_name = "pytest"
+    yield params_tbl, params_name
+
+
+@pytest.fixture(scope="session")
+def training_params_key(verbose_context, sgp, project_key, dlc_training_params):
+    params_tbl, params_name = dlc_training_params
     with verbose_context:
-        sgp.v1.DLCModelTrainingParams.insert_new_params(
-            paramset_name=training_params_name,
+        params_tbl.insert_new_params(
+            paramset_name=params_name,
             params={
                 "trainingsetindex": 0,
                 "shuffle": 1,
@@ -901,10 +928,11 @@ def training_params_key(verbose_context, sgp, project_key):
                 "TFGPUinference": False,
                 "net_type": "resnet_50",
                 "augmenter_type": "imgaug",
+                "video_sets": "test skipping param",
             },
             skip_duplicates=True,
         )
-    yield {"dlc_training_params_name": training_params_name}
+    yield {"dlc_training_params_name": params_name}
 
 
 @pytest.fixture(scope="session")
@@ -913,7 +941,6 @@ def model_train_key(sgp, project_key, training_params_key):
     model_train_key = {
         **project_key,
         **training_params_key,
-        "training_id": 0,
     }
     sgp.v1.DLCModelTrainingSelection().insert1(
         {
@@ -974,19 +1001,17 @@ def pose_estimation_key(sgp, mini_copy_name, populate_model, model_key):
 
 @pytest.fixture(scope="session")
 def populate_pose_estimation(sgp, pose_estimation_key):
-    pose_est_tbl = sgp.v1.DLCPoseEstimation
-    if pose_est_tbl & pose_estimation_key:
-        yield
-    else:
+    pose_est_tbl = sgp.v1.DLCPoseEstimation()
+    if len(pose_est_tbl & pose_estimation_key) < 1:
         pose_est_tbl.populate(pose_estimation_key)
-        yield
+    yield pose_est_tbl
 
 
 @pytest.fixture(scope="session")
 def si_params_name(sgp, populate_pose_estimation):
     params_name = "low_bar"
     params_tbl = sgp.v1.DLCSmoothInterpParams
-    # if len(params_tbl & {"dlc_si_params_name": params_name}) == 0:
+    # if len(params_tbl & {"dlc_si_params_name": params_name}) < 1:
     if True:  # TODO: remove before merge
         nan_params = params_tbl.get_nan_params()
         nan_params["dlc_si_params_name"] = params_name
@@ -995,6 +1020,9 @@ def si_params_name(sgp, populate_pose_estimation):
                 "likelihood_thresh": 0.4,
                 "max_cm_between_pts": 100,
                 "num_inds_to_span": 50,
+                # Smoothing and Interpolation added later - must check
+                "smoothing_params": {"smoothing_duration": 0.05},
+                "interp_params": {"max_cm_to_interp": 100},
             }
         )
         params_tbl.insert1(nan_params, skip_duplicates=True)
diff --git a/tests/position/conftest.py b/tests/position/conftest.py
index 1aaec3384..caf88448a 100644
--- a/tests/position/conftest.py
+++ b/tests/position/conftest.py
@@ -11,15 +11,6 @@
     58, 61, 69, 72, 97-100, 104, 149-161, 232-235, 239-241, 246, 259, 280,
     293-305, 310-316, 328-341, 356-373, 395, 404, 480, 487-488, 530, 548-561,
     594-601, 611-612, 641-657, 682-736, 762-772, 787, 809-1286
-
-TODO: tests for
-pose_estimat   51-71, 102, 115, 256, 345-366
-position.py    53, 99, 114, 119, 197-198, 205-219, 349, 353-355, 360, 382, 385, 407, 443-466
-project.py     45-54, 128-205, 250-255, 259, 278-303, 316, 347, 361-413, 425, 457, 476-479, 486-489, 514-555, 582, 596
-selection.py   213, 282, 308-417
-training.py    55, 67-73, 85-87, 113, 143-144, 161, 207-210
-es_position.py 67, 282-283, 361-362, 496, 502-503
-
 """
 
 from itertools import product as iter_prodect
@@ -33,7 +24,7 @@
 def dlc_video_params(sgp):
     sgp.v1.DLCPosVideoParams.insert_default()
     params_key = {"dlc_pos_video_params_name": "five_percent"}
-    sgp.v1.DLCPosVideoSelection.insert1(
+    sgp.v1.DLCPosVideoParams.insert1(
         {
             **params_key,
             "params": {
@@ -47,7 +38,7 @@ def dlc_video_params(sgp):
 
 
 @pytest.fixture(scope="session")
-def dlc_video_selection(sgp, dlc_key, dlc_video_params):
+def dlc_video_selection(sgp, dlc_key, dlc_video_params, populate_dlc):
     s_key = {**dlc_key, **dlc_video_params}
     sgp.v1.DLCPosVideoSelection.insert1(s_key, skip_duplicates=True)
     yield dlc_key
@@ -56,7 +47,7 @@ def dlc_video_selection(sgp, dlc_key, dlc_video_params):
 @pytest.fixture(scope="session")
 def populate_dlc_video(sgp, dlc_video_selection):
     sgp.v1.DLCPosVideo.populate(dlc_video_selection)
-    yield
+    yield sgp.v1.DLCPosVideo()
 
 
 @pytest.fixture(scope="session")
@@ -95,4 +86,4 @@ def increment_count():
         count[0] += 1
         return count[0]
 
-    return df.map(lambda x: increment_count() if x == 1 else x)
+    return df.applymap(lambda x: increment_count() if x == 1 else x)
diff --git a/tests/position/test_dlc_cent.py b/tests/position/test_dlc_cent.py
index 86ccba275..b312ebff5 100644
--- a/tests/position/test_dlc_cent.py
+++ b/tests/position/test_dlc_cent.py
@@ -10,22 +10,11 @@ def centroid_df(sgp, centroid_key, populate_centroid):
     yield (sgp.v1.DLCCentroid & centroid_key).fetch1_dataframe()
 
 
-@pytest.mark.parametrize(
-    "column, exp_sum",
-    [
-        ("video_frame_ind", 36312),
-        ("position_x", 17987),
-        ("position_y", 2983),
-        ("velocity_x", -1.489),
-        ("velocity_y", 4.160),
-        ("speed", 12957),
-    ],
-)
-def test_centroid_fetch1_dataframe(centroid_df, column, exp_sum):
-    tolerance = abs(centroid_df[column].iloc[0] * 0.1)
+def test_centroid_fetch1_dataframe(centroid_df):
+    df_sum = centroid_df.sum().sum()
     assert np_isclose(
-        centroid_df[column].sum(), exp_sum, atol=tolerance
-    ), f"Sum of {column} in Centroid dataframe is not as expected"
+        df_sum, 55_860, atol=1000
+    ), f"Unexpected checksum for centroid dataframe: {df_sum}"
 
 
 @pytest.fixture(scope="session")
@@ -45,8 +34,8 @@ def test_insert_default_params(params_tbl):
 
 def test_validate_params(params_tbl):
     params = params_tbl.get_default()
-    params["dlc_centroid_params_name"] = "test"
-    params_tbl.insert1(params)
+    params["dlc_centroid_params_name"] = "other test"
+    params_tbl.insert1(params, skip_duplicates=True)
 
 
 @pytest.mark.parametrize(
diff --git a/tests/position/test_dlc_pos.py b/tests/position/test_dlc_pos.py
deleted file mode 100644
index df878c90c..000000000
--- a/tests/position/test_dlc_pos.py
+++ /dev/null
@@ -1,52 +0,0 @@
-import pytest
-from numpy import isclose as np_isclose
-
-
-def test_si_params_default(sgp):
-    assert sgp.v1.DLCSmoothInterpParams.get_default() == {
-        "dlc_si_params_name": "default",
-        "params": {
-            "interp_params": {"max_cm_to_interp": 15},
-            "interpolate": True,
-            "likelihood_thresh": 0.95,
-            "max_cm_between_pts": 20,
-            "num_inds_to_span": 20,
-            "smooth": True,
-            "smoothing_params": {
-                "smooth_method": "moving_avg",
-                "smoothing_duration": 0.05,
-            },
-        },
-    }
-    assert sgp.v1.DLCSmoothInterpParams.get_nan_params() == {
-        "dlc_si_params_name": "just_nan",
-        "params": {
-            "interpolate": False,
-            "likelihood_thresh": 0.95,
-            "max_cm_between_pts": 20,
-            "num_inds_to_span": 20,
-            "smooth": False,
-        },
-    }
-
-
-@pytest.fixture(scope="session")
-def si_df(sgp, si_key, populate_si, bodyparts):
-    yield (
-        sgp.v1.DLCSmoothInterp() & {**si_key, "bodypart": bodyparts[0]}
-    ).fetch1_dataframe()
-
-
-@pytest.mark.parametrize(
-    "column, exp_sum",
-    [
-        ("video_frame_ind", 36312),
-        ("x", 17987),
-        ("y", 2983),
-    ],
-)
-def test_centroid_fetch1_dataframe(si_df, column, exp_sum):
-    tolerance = abs(si_df[column].iloc[0] * 0.1)
-    assert np_isclose(
-        si_df[column].sum(), exp_sum, atol=tolerance
-    ), f"Sum of {column} in SmoothInterp dataframe is not as expected"
diff --git a/tests/position/test_dlc_pos_est.py b/tests/position/test_dlc_pos_est.py
new file mode 100644
index 000000000..e011664ad
--- /dev/null
+++ b/tests/position/test_dlc_pos_est.py
@@ -0,0 +1,34 @@
+import pytest
+
+
+@pytest.fixture(scope="session")
+def pos_est_sel(sgp):
+    yield sgp.v1.position_dlc_pose_estimation.DLCPoseEstimationSelection()
+
+
+def test_rename_non_default_columns(sgp, common, pos_est_sel, video_keys):
+    vid_path, vid_name, _, _ = sgp.v1.dlc_utils.get_video_path(video_keys[0])
+
+    input = "0, 10, 0, 1000"
+    output = pos_est_sel.get_video_crop(vid_path + vid_name, input)
+
+    assert (
+        input == output
+    ), f"{pos_est_sel.table_name}.get_video_crop did not return expected output"
+
+
+def test_invalid_video(pos_est_sel, pose_estimation_key):
+    _ = pose_estimation_key  # Ensure populated
+    example_key = pos_est_sel.fetch("KEY", as_dict=True)[0]
+    example_key["nwb_file_name"] = "invalid.nwb"
+    with pytest.raises(FileNotFoundError):
+        pos_est_sel.insert_estimation_task(example_key)
+
+
+def test_pose_est_dataframe(populate_pose_estimation):
+    pose_cols = populate_pose_estimation.fetch_dataframe().columns
+
+    for bp in ["tailBase", "tailMid", "tailTip"]:
+        for val in ["video_frame_ind", "x", "y"]:
+            col = (bp, val)
+            assert col in pose_cols, f"PoseEstimation df missing column {col}."
diff --git a/tests/position/test_dlc_position.py b/tests/position/test_dlc_position.py
new file mode 100644
index 000000000..94646f315
--- /dev/null
+++ b/tests/position/test_dlc_position.py
@@ -0,0 +1,64 @@
+import pytest
+
+
+@pytest.fixture(scope="session")
+def si_params_tbl(sgp):
+    yield sgp.v1.DLCSmoothInterpParams()
+
+
+def test_si_params_default(si_params_tbl):
+    assert si_params_tbl.get_default() == {
+        "dlc_si_params_name": "default",
+        "params": {
+            "interp_params": {"max_cm_to_interp": 15},
+            "interpolate": True,
+            "likelihood_thresh": 0.95,
+            "max_cm_between_pts": 20,
+            "num_inds_to_span": 20,
+            "smooth": True,
+            "smoothing_params": {
+                "smooth_method": "moving_avg",
+                "smoothing_duration": 0.05,
+            },
+        },
+    }
+    assert si_params_tbl.get_nan_params() == {
+        "dlc_si_params_name": "just_nan",
+        "params": {
+            "interpolate": False,
+            "likelihood_thresh": 0.95,
+            "max_cm_between_pts": 20,
+            "num_inds_to_span": 20,
+            "smooth": False,
+        },
+    }
+    assert list(si_params_tbl.get_available_methods()) == [
+        "moving_avg"
+    ], f"{si_params_tbl.table_name}: unexpected available methods"
+
+
+def test_invalid_params_insert(si_params_tbl):
+    with pytest.raises(KeyError):
+        si_params_tbl.insert1({"params": "invalid"})
+
+
+@pytest.fixture(scope="session")
+def si_df(sgp, si_key, populate_si, bodyparts):
+    yield (
+        sgp.v1.DLCSmoothInterp() & {**si_key, "bodypart": bodyparts[0]}
+    ).fetch1_dataframe()
+
+
+def test_cohort_fetch1_dataframe(si_df):
+    df_cols = si_df.columns
+    exp_cols = ["video_frame_ind", "x", "y"]
+    assert all(
+        e in df_cols for e in exp_cols
+    ), f"Unexpected cols in DLCSmoothInterp dataframe: {df_cols}"
+
+
+def test_all_nans(populate_pose_estimation, sgp):
+    pose_est_tbl = populate_pose_estimation
+    df = pose_est_tbl.BodyPart().fetch1_dataframe()
+    with pytest.raises(ValueError):
+        sgp.v1.position_dlc_position.nan_inds(df, 10, 0.99, 10)
diff --git a/tests/position/test_dlc_proj.py b/tests/position/test_dlc_proj.py
index d3236b1c5..7eaba196d 100644
--- a/tests/position/test_dlc_proj.py
+++ b/tests/position/test_dlc_proj.py
@@ -1,7 +1,68 @@
+import pytest
+
+
+def test_bp_insert(sgp):
+    bp_tbl = sgp.v1.position_dlc_project.BodyPart()
+
+    bp_w_desc, desc = "test_bp", "test_desc"
+    bp_no_desc = "test_bp_no_desc"
+
+    bp_tbl.add_from_config([bp_w_desc], [desc])
+    bp_tbl.add_from_config([bp_no_desc])
+
+    assert bp_tbl & {
+        "bodypart": bp_w_desc,
+        "description": desc,
+    }, "Bodypart with description not inserted correctly"
+    assert bp_tbl & {
+        "bodypart": bp_no_desc,
+        "description": bp_no_desc,
+    }, "Bodypart without description not inserted correctly"
+
+
 def test_project_insert(dlc_project_tbl, project_key):
     assert dlc_project_tbl & project_key, "Project not inserted correctly"
 
 
+@pytest.fixture
+def new_project_key():
+    return {
+        "project_name": "test_project_name",
+        "bodyparts": ["bp1"],
+        "lab_team": "any",
+        "frames_per_video": 1,
+        "video_list": ["any"],
+        "groupname": "fake group",
+    }
+
+
+def test_failed_name_insert(
+    dlc_project_tbl, dlc_project_name, config_path, new_project_key
+):
+    new_project_key.update({"project_name": dlc_project_name})
+    existing_key = dlc_project_tbl.insert_new_project(
+        project_name=dlc_project_name,
+        bodyparts=["bp1"],
+        lab_team="any",
+        frames_per_video=1,
+        video_list=["any"],
+        groupname="any",
+    )
+    expected_key = {
+        "project_name": dlc_project_name,
+        "config_path": config_path,
+    }
+    assert (
+        existing_key == expected_key
+    ), "Project re-insert did not return expected key"
+
+
+def test_failed_group_insert(dlc_project_tbl, new_project_key):
+    with pytest.raises(ValueError):
+        dlc_project_tbl.insert_new_project(**new_project_key)
+
+
 def test_extract_frames(extract_frames, labeled_vid_dir):
     extracted_files = list(labeled_vid_dir.glob("*.png"))
-    assert len(extracted_files) == 4, "Incorrect number of frames extracted"
+    stems = set([f.stem for f in extracted_files]) - {"img000", "img001"}
+    assert len(stems) == 2, "Incorrect number of frames extracted"
diff --git a/tests/position/test_dlc_sel.py b/tests/position/test_dlc_sel.py
index 9d869f8bb..35b33fe06 100644
--- a/tests/position/test_dlc_sel.py
+++ b/tests/position/test_dlc_sel.py
@@ -1,5 +1,5 @@
 def test_dlcvideo_default(sgp):
-    assert sgp.v1.DLCPosVideoParams.get_default() == {
+    expected_default = {
         "dlc_pos_video_params_name": "default",
         "params": {
             "incl_likelihood": True,
@@ -7,3 +7,11 @@ def test_dlcvideo_default(sgp):
             "video_params": {"arrow_radius": 20, "circle_radius": 6},
         },
     }
+
+    # run twice to trigger fetch existing
+    assert sgp.v1.DLCPosVideoParams.get_default() == expected_default
+    assert sgp.v1.DLCPosVideoParams.get_default() == expected_default
+
+
+def test_dlc_video_populate(populate_dlc_video):
+    assert len(populate_dlc_video) > 0, "DLCPosVideo table is empty"
diff --git a/tests/position/test_dlc_train.py b/tests/position/test_dlc_train.py
new file mode 100644
index 000000000..58204abb4
--- /dev/null
+++ b/tests/position/test_dlc_train.py
@@ -0,0 +1,30 @@
+def test_existing_params(
+    verbose_context, dlc_training_params, training_params_key
+):
+    params_tbl, params_name = dlc_training_params
+
+    _ = training_params_key  # Ensure populated
+    params_query = params_tbl & {"dlc_training_params_name": params_name}
+    assert params_query, "Existing params not found"
+
+    with verbose_context:
+        params_tbl.insert_new_params(
+            paramset_name=params_name,
+            params={
+                "shuffle": 1,
+                "trainingsetindex": 0,
+                "net_type": "any",
+                "gputouse": None,
+            },
+            skip_duplicates=False,
+        )
+
+    assert len(params_query) == 1, "Existing params duplicated"
+
+
+def test_get_params(verbose_context, dlc_training_params):
+    params_tbl, _ = dlc_training_params
+    with verbose_context:
+        accepted_params = params_tbl.get_accepted_params()
+
+    assert accepted_params is not None, "Failed to get accepted params"
diff --git a/tests/position/test_pos_merge.py b/tests/position/test_pos_merge.py
index af6b17e0f..047129cd5 100644
--- a/tests/position/test_pos_merge.py
+++ b/tests/position/test_pos_merge.py
@@ -1,5 +1,4 @@
 import pytest
-from numpy import isclose as np_isclose
 
 
 @pytest.fixture(scope="session")
@@ -8,19 +7,18 @@ def merge_df(sgp, pos_merge, dlc_key, populate_dlc):
     yield (pos_merge & merge_key).fetch1_dataframe()
 
 
-@pytest.mark.parametrize(
-    "column, exp_sum",
-    [  # NOTE: same as test_centroid_fetch1_dataframe
-        ("video_frame_ind", 36312),
-        ("position_x", 17987),
-        ("position_y", 2983),
-        ("velocity_x", -1.489),
-        ("velocity_y", 4.160),
-        ("speed", 12957),
-    ],
-)
-def test_merge_dlc_fetch1_dataframe(merge_df, column, exp_sum):
-    tolerance = abs(merge_df[column].iloc[0] * 0.1)
-    assert np_isclose(
-        merge_df[column].sum(), exp_sum, atol=tolerance
-    ), f"Sum of {column} in Merge.DLCPosV1 dataframe is not as expected"
+def test_merge_dlc_fetch1_dataframe(merge_df):
+    df_cols = merge_df.columns
+    exp_cols = [
+        "video_frame_ind",
+        "position_x",
+        "position_y",
+        "orientation",
+        "velocity_x",
+        "velocity_y",
+        "speed",
+    ]
+
+    assert all(
+        e in df_cols for e in exp_cols
+    ), f"Unexpected cols in position merge dataframe: {df_cols}"