atomscale 0.19.0__tar.gz → 0.20.0__tar.gz
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- {atomscale-0.19.0/src/atomscale.egg-info → atomscale-0.20.0}/PKG-INFO +1 -1
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/client.py +173 -0
- atomscale-0.20.0/src/atomscale/rheed_metadata.py +127 -0
- atomscale-0.20.0/src/atomscale/samples.py +292 -0
- {atomscale-0.19.0 → atomscale-0.20.0/src/atomscale.egg-info}/PKG-INFO +1 -1
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale.egg-info/SOURCES.txt +3 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale.egg-info/scm_file_list.json +3 -0
- atomscale-0.20.0/src/atomscale.egg-info/scm_version.json +8 -0
- atomscale-0.20.0/tests/test_samples.py +421 -0
- atomscale-0.19.0/src/atomscale.egg-info/scm_version.json +0 -8
- {atomscale-0.19.0 → atomscale-0.20.0}/.github/workflows/release.yml +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/.github/workflows/testing.yml +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/.github/workflows/upgrade_dependencies.yml +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/.gitignore +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/.pre-commit-config.yaml +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/CHANGELOG.md +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/LICENSE +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/MANIFEST.in +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/README.md +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/atomicds-shim-dist/pyproject.toml +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/docs/Makefile +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/docs/_static/AtomscaleLogoFull.png +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/docs/_static/custom.css +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/docs/_templates/custom-class-template.rst +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/docs/_templates/custom-module-template.rst +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/docs/conf.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/docs/guides/analysis-results.rst +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/docs/guides/monitor-live.rst +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/docs/guides/quickstart.rst +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/docs/guides/search-and-download.rst +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/docs/guides/stream-rheed.rst +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/docs/guides/stream-timeseries.rst +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/docs/guides/upload-files.rst +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/docs/index.rst +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/docs/make.bat +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/docs/modules.rst +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/examples/changepoints.ipynb +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/examples/general_use.ipynb +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/examples/rheed_streaming.ipynb +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/examples/timeseries_polling.ipynb +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/examples/vxwse2-placeholder/task1_films.ipynb +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/examples/vxwse2-placeholder/task1_sapphire.ipynb +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/examples/vxwse2-placeholder/task2_composition.ipynb +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/pyproject.toml +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-macos-latest_py3.10.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-macos-latest_py3.10_extras.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-macos-latest_py3.11.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-macos-latest_py3.11_extras.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-macos-latest_py3.12.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-macos-latest_py3.12_extras.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-macos-latest_py3.13.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-macos-latest_py3.13_extras.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-ubuntu-latest_py3.10.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-ubuntu-latest_py3.10_extras.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-ubuntu-latest_py3.11.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-ubuntu-latest_py3.11_extras.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-ubuntu-latest_py3.12.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-ubuntu-latest_py3.12_extras.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-ubuntu-latest_py3.13.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-ubuntu-latest_py3.13_extras.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-windows-latest_py3.10.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-windows-latest_py3.10_extras.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-windows-latest_py3.11.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-windows-latest_py3.11_extras.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-windows-latest_py3.12.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-windows-latest_py3.12_extras.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-windows-latest_py3.13.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/requirements/requirements-windows-latest_py3.13_extras.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/scripts/ci_auth_preflight.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/setup.cfg +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomicds/__init__.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/__init__.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/core/__init__.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/core/client.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/core/files.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/core/utils.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/results/__init__.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/results/changepoint.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/results/ellipsometry.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/results/embeddings.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/results/group.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/results/metrology.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/results/optical.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/results/photoluminescence.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/results/raman.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/results/recipe.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/results/rheed_image.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/results/rheed_video.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/results/similarity_trajectory.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/results/tool_state.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/results/unknown.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/results/xps.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/results/xrd.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/similarity/__init__.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/similarity/embedding_provider.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/similarity/polling.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/similarity/provider.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/streaming/Cargo.lock +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/streaming/Cargo.toml +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/streaming/__init__.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/streaming/rheed_stream.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/streaming/rheed_stream.pyi +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/streaming/src/initialize.rs +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/streaming/src/lib.rs +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/streaming/src/timeseries.rs +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/streaming/src/upload.rs +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/streaming/src/utils.rs +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/timeseries/__init__.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/timeseries/_polling_utils.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/timeseries/align.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/timeseries/ellipsometry.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/timeseries/metrology.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/timeseries/optical.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/timeseries/physical_sample.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/timeseries/polling.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/timeseries/provider.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/timeseries/recipe.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/timeseries/registry.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/timeseries/rheed.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/timeseries/sample.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale/timeseries/tool_state.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale.egg-info/dependency_links.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale.egg-info/requires.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/src/atomscale.egg-info/top_level.txt +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/__init__.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/_mock_http_server.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/conftest.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/data/test_rheed.mp4 +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_align.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_atomicds_alias.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_changepoint.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_client.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_core.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_ellipsometry.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_embeddings.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_metrology.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_optical.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_photoluminescence.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_physical_sample_timeseries.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_polling.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_raman.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_recipe.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_rheed_embedding.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_rheed_image.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_rheed_stream.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_rheed_timeseries.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_rheed_video.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_similarity_matches.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_similarity_trajectory.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_timeseries_stream.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/tests/test_xps.py +0 -0
- {atomscale-0.19.0 → atomscale-0.20.0}/uv.lock +0 -0
|
@@ -40,6 +40,13 @@ from atomscale.results import (
|
|
|
40
40
|
decode_mask_rle,
|
|
41
41
|
)
|
|
42
42
|
from atomscale.results.group import PhysicalSampleResult, ProjectResult
|
|
43
|
+
from atomscale.rheed_metadata import rheed_azimuths_to_dataframe
|
|
44
|
+
from atomscale.samples import (
|
|
45
|
+
ai_summaries_to_dataframe,
|
|
46
|
+
ai_summary_row,
|
|
47
|
+
process_steps_to_dataframe,
|
|
48
|
+
spatial_annotations_to_dataframe,
|
|
49
|
+
)
|
|
43
50
|
from atomscale.timeseries.align import align_timeseries
|
|
44
51
|
from atomscale.timeseries.physical_sample import (
|
|
45
52
|
physical_sample_timeseries_to_dataframe,
|
|
@@ -720,6 +727,39 @@ class Client(BaseClient):
|
|
|
720
727
|
ts_df = provider.attach_frame_masks(ts_df, mask_rows) # type: ignore[arg-type]
|
|
721
728
|
return ts_df
|
|
722
729
|
|
|
730
|
+
def get_rheed_azimuths(self, data_ids: str | list[str]) -> DataFrame:
|
|
731
|
+
"""Per-azimuth metadata for rotating RHEED recordings.
|
|
732
|
+
|
|
733
|
+
For each seed frame of a rotating video this returns both the rotation
|
|
734
|
+
angle it sits at and the crystallographic azimuth the production
|
|
735
|
+
classifier assigned it. The angle depends on where the substrate happened
|
|
736
|
+
to be parked, so the same azimuth appears at different angles in different
|
|
737
|
+
recordings of one sample; the label is the stable identity, and so the
|
|
738
|
+
right key for aligning or concatenating series across recordings.
|
|
739
|
+
|
|
740
|
+
Args:
|
|
741
|
+
data_ids: Data ID or list of data IDs of rotating RHEED videos.
|
|
742
|
+
|
|
743
|
+
Returns:
|
|
744
|
+
DataFrame: One row per (data item, seed frame) with columns
|
|
745
|
+
``data_id``, ``seed_frame``, ``angle_degrees``, ``azimuth_label``
|
|
746
|
+
(``"100"`` / ``"110"`` / ``"210"``, or ``None`` when the classifier
|
|
747
|
+
could not call it), ``label_confidence``, ``crystal_system`` and
|
|
748
|
+
``surface_miller``. Data items with no rotating configuration
|
|
749
|
+
contribute no rows.
|
|
750
|
+
"""
|
|
751
|
+
if isinstance(data_ids, str):
|
|
752
|
+
data_ids = [data_ids]
|
|
753
|
+
|
|
754
|
+
payload = self._get(
|
|
755
|
+
sub_url="configuration/rheed/video/", params={"data_ids": data_ids}
|
|
756
|
+
)
|
|
757
|
+
if payload is None:
|
|
758
|
+
return rheed_azimuths_to_dataframe([])
|
|
759
|
+
return rheed_azimuths_to_dataframe(
|
|
760
|
+
payload if isinstance(payload, list) else [payload]
|
|
761
|
+
)
|
|
762
|
+
|
|
723
763
|
def get_frame(
|
|
724
764
|
self,
|
|
725
765
|
data_id: str,
|
|
@@ -1207,6 +1247,139 @@ class Client(BaseClient):
|
|
|
1207
1247
|
raw, property_names=property_names
|
|
1208
1248
|
)
|
|
1209
1249
|
|
|
1250
|
+
def get_process_steps(self, physical_sample_id: str) -> DataFrame:
|
|
1251
|
+
"""The physical sample's process steps — its temporal phases.
|
|
1252
|
+
|
|
1253
|
+
A step groups the data items whose acquisition windows cluster together,
|
|
1254
|
+
so one sample's run splits into the phases it actually went through
|
|
1255
|
+
(e.g. a substrate anneal, then a growth, then a cooldown). Steps are the
|
|
1256
|
+
natural scope for a per-run report: the newest step is "this run", and
|
|
1257
|
+
the earlier ones are its history.
|
|
1258
|
+
|
|
1259
|
+
Args:
|
|
1260
|
+
physical_sample_id: Identifier of the physical sample.
|
|
1261
|
+
|
|
1262
|
+
Returns:
|
|
1263
|
+
DataFrame: One row per step ordered by ``step_index``, with columns
|
|
1264
|
+
``step_index``, ``process_step_id``, ``name`` (the human/material
|
|
1265
|
+
label, ``None`` until set), ``start_unix_ms_utc`` /
|
|
1266
|
+
``end_unix_ms_utc``, ``duration_seconds``, the tz-aware
|
|
1267
|
+
``start_datetime`` / ``end_datetime``, ``n_data_ids``, ``data_ids``
|
|
1268
|
+
(tuple of the step's data items), and ``last_updated``. A sample
|
|
1269
|
+
whose steps have never been computed yields an empty DataFrame.
|
|
1270
|
+
|
|
1271
|
+
Raises:
|
|
1272
|
+
ClientError: If the physical sample does not exist (404).
|
|
1273
|
+
"""
|
|
1274
|
+
raw = self._get(sub_url=f"physical_samples/{physical_sample_id}/process_steps")
|
|
1275
|
+
if raw is None:
|
|
1276
|
+
raise ClientError(
|
|
1277
|
+
f"No physical sample found with id '{physical_sample_id}' "
|
|
1278
|
+
"(physical_samples/{id}/process_steps returned 404).",
|
|
1279
|
+
status_code=404,
|
|
1280
|
+
)
|
|
1281
|
+
return process_steps_to_dataframe(raw) # type: ignore[arg-type]
|
|
1282
|
+
|
|
1283
|
+
def get_spatial_annotations(
|
|
1284
|
+
self,
|
|
1285
|
+
physical_sample_id: str,
|
|
1286
|
+
*,
|
|
1287
|
+
property_name: str | None = None,
|
|
1288
|
+
char_source_type: str | None = None,
|
|
1289
|
+
x_range: tuple[float | None, float | None] = (None, None),
|
|
1290
|
+
y_range: tuple[float | None, float | None] = (None, None),
|
|
1291
|
+
) -> DataFrame:
|
|
1292
|
+
"""Spatially-resolved property values pinned to a sample's surface.
|
|
1293
|
+
|
|
1294
|
+
Each annotation is one property measured at one position on the sample —
|
|
1295
|
+
an XPS composition ratio stepped across a wafer, a thickness from a
|
|
1296
|
+
mapped ellipsometry scan — which is what a spatial (wafer-map) view of a
|
|
1297
|
+
run is built from.
|
|
1298
|
+
|
|
1299
|
+
Args:
|
|
1300
|
+
physical_sample_id: Identifier of the physical sample.
|
|
1301
|
+
property_name: Restrict to a single property, e.g. ``"sr_ti_ratio"``.
|
|
1302
|
+
char_source_type: Restrict to the technique that produced the values,
|
|
1303
|
+
e.g. ``"xps"``.
|
|
1304
|
+
x_range: Inclusive ``(min, max)`` bounding-box filter on ``coord_x``.
|
|
1305
|
+
The endpoint applies a coordinate filter only when *both* bounds
|
|
1306
|
+
are given, so a half-open range is ignored server-side.
|
|
1307
|
+
y_range: Inclusive ``(min, max)`` bounding-box filter on ``coord_y``,
|
|
1308
|
+
with the same both-bounds requirement as ``x_range``.
|
|
1309
|
+
|
|
1310
|
+
Returns:
|
|
1311
|
+
DataFrame: One row per annotated point with columns
|
|
1312
|
+
``annotation_id``, ``data_id``, ``char_source_type``,
|
|
1313
|
+
``property_name``, ``property_value``, ``property_unit``,
|
|
1314
|
+
``coord_x`` / ``coord_y`` / ``coord_z``, ``coord_ref_frame``,
|
|
1315
|
+
``coord_units``, ``metadata``, and ``last_updated``. Sorted by
|
|
1316
|
+
property then position. A sample with no annotations — or a filter
|
|
1317
|
+
that matches none — yields an empty DataFrame.
|
|
1318
|
+
|
|
1319
|
+
Raises:
|
|
1320
|
+
ClientError: If the physical sample does not exist (404).
|
|
1321
|
+
"""
|
|
1322
|
+
raw = self._get(
|
|
1323
|
+
sub_url=f"physical_samples/{physical_sample_id}/annotations/",
|
|
1324
|
+
params={
|
|
1325
|
+
"property_name": property_name,
|
|
1326
|
+
"char_source_type": char_source_type,
|
|
1327
|
+
"x_min": x_range[0],
|
|
1328
|
+
"x_max": x_range[1],
|
|
1329
|
+
"y_min": y_range[0],
|
|
1330
|
+
"y_max": y_range[1],
|
|
1331
|
+
},
|
|
1332
|
+
)
|
|
1333
|
+
# An existing sample with no annotations returns ``[]``, which ``_get``
|
|
1334
|
+
# hands back as an empty list; only a missing sample 404s to ``None``.
|
|
1335
|
+
if raw is None:
|
|
1336
|
+
raise ClientError(
|
|
1337
|
+
f"No physical sample found with id '{physical_sample_id}' "
|
|
1338
|
+
"(physical_samples/{id}/annotations/ returned 404).",
|
|
1339
|
+
status_code=404,
|
|
1340
|
+
)
|
|
1341
|
+
return spatial_annotations_to_dataframe(raw) # type: ignore[arg-type]
|
|
1342
|
+
|
|
1343
|
+
def get_ai_summaries(self, data_ids: str | list[str]) -> DataFrame:
|
|
1344
|
+
"""Agent-written summaries for one or more data items.
|
|
1345
|
+
|
|
1346
|
+
The summary is generated asynchronously, so every requested data id gets
|
|
1347
|
+
a row carrying its ``task_status``: ``"available"`` when a summary
|
|
1348
|
+
exists, ``"pending"`` while the summarization workflow is still running,
|
|
1349
|
+
and ``"not_found"`` when neither is true. That makes an un-summarized
|
|
1350
|
+
item visible to a caller rendering a report rather than silently absent.
|
|
1351
|
+
|
|
1352
|
+
Args:
|
|
1353
|
+
data_ids: Data ID or list of data IDs from the data catalogue.
|
|
1354
|
+
|
|
1355
|
+
Returns:
|
|
1356
|
+
DataFrame: One row per requested data id, in request order, with
|
|
1357
|
+
columns ``data_id``, ``task_status``, ``summary_id``, ``content``
|
|
1358
|
+
(the prose summary), ``structured_content`` (the agent's full
|
|
1359
|
+
structured output), ``generation_model``, ``generation_timestamp``,
|
|
1360
|
+
``user_feedback``, and ``workflow_id``. Summary-note columns are
|
|
1361
|
+
``None`` for rows whose status is not ``"available"``.
|
|
1362
|
+
"""
|
|
1363
|
+
if isinstance(data_ids, str):
|
|
1364
|
+
data_ids = [data_ids]
|
|
1365
|
+
|
|
1366
|
+
# The endpoint is per-data-item, so fan the requests out rather than
|
|
1367
|
+
# walking them serially.
|
|
1368
|
+
with _make_progress(self.mute_bars, False) as progress:
|
|
1369
|
+
payloads = self._multi_thread(
|
|
1370
|
+
lambda data_id: self._get(sub_url=f"summary/{data_id}"),
|
|
1371
|
+
[{"data_id": data_id} for data_id in data_ids],
|
|
1372
|
+
progress,
|
|
1373
|
+
progress_description="Obtaining AI summaries",
|
|
1374
|
+
)
|
|
1375
|
+
|
|
1376
|
+
return ai_summaries_to_dataframe(
|
|
1377
|
+
[
|
|
1378
|
+
ai_summary_row(data_id, payload)
|
|
1379
|
+
for data_id, payload in zip(data_ids, payloads, strict=True)
|
|
1380
|
+
]
|
|
1381
|
+
)
|
|
1382
|
+
|
|
1210
1383
|
def get_project(
|
|
1211
1384
|
self,
|
|
1212
1385
|
project_id: str,
|
|
@@ -0,0 +1,127 @@
|
|
|
1
|
+
"""Parse per-azimuth RHEED metadata for a rotating-substrate recording.
|
|
2
|
+
|
|
3
|
+
``GET /configuration/rheed/video/`` carries, for each seed frame of a rotating
|
|
4
|
+
video, both the rotation angle the frame sits at (``seed_frame_angles``) and the
|
|
5
|
+
crystallographic azimuth the production classifier assigned it
|
|
6
|
+
(``api_configuration.azimuth_labels`` — one of ``"100"`` / ``"110"`` / ``"210"``,
|
|
7
|
+
or ``None`` when it could not be called), plus the fit detail behind that call in
|
|
8
|
+
``azimuth_label_meta``.
|
|
9
|
+
|
|
10
|
+
The distinction matters when comparing recordings: the rotation angle is a
|
|
11
|
+
property of how the substrate happened to be parked, so the *same* azimuth turns
|
|
12
|
+
up at different angles in different recordings of the same sample. The
|
|
13
|
+
crystallographic label is the stable identity, and therefore the correct key for
|
|
14
|
+
aligning or concatenating series across recordings.
|
|
15
|
+
"""
|
|
16
|
+
|
|
17
|
+
from __future__ import annotations
|
|
18
|
+
|
|
19
|
+
from collections.abc import Mapping, Sequence
|
|
20
|
+
from typing import Any
|
|
21
|
+
|
|
22
|
+
import pandas as pd
|
|
23
|
+
from pandas import DataFrame
|
|
24
|
+
|
|
25
|
+
# One row per (data item, seed frame).
|
|
26
|
+
RHEED_AZIMUTH_COLUMNS: tuple[str, ...] = (
|
|
27
|
+
"data_id",
|
|
28
|
+
"seed_frame",
|
|
29
|
+
"angle_degrees",
|
|
30
|
+
"azimuth_label",
|
|
31
|
+
"label_confidence",
|
|
32
|
+
"crystal_system",
|
|
33
|
+
"surface_miller",
|
|
34
|
+
)
|
|
35
|
+
|
|
36
|
+
|
|
37
|
+
def _empty() -> DataFrame:
|
|
38
|
+
frame = DataFrame({c: [] for c in RHEED_AZIMUTH_COLUMNS})
|
|
39
|
+
frame["seed_frame"] = frame["seed_frame"].astype("int64")
|
|
40
|
+
frame["angle_degrees"] = frame["angle_degrees"].astype("float64")
|
|
41
|
+
frame["label_confidence"] = frame["label_confidence"].astype("float64")
|
|
42
|
+
return frame
|
|
43
|
+
|
|
44
|
+
|
|
45
|
+
def rheed_azimuths_to_dataframe(payload: Sequence[Mapping[str, Any]]) -> DataFrame:
|
|
46
|
+
"""Convert a ``list[RheedVideoConfigurationResponse]`` to one row per azimuth.
|
|
47
|
+
|
|
48
|
+
Args:
|
|
49
|
+
payload: Decoded response body — one entry per requested data id, each
|
|
50
|
+
with ``data_id``, ``seed_frame_angles`` (seed frame -> degrees) and an
|
|
51
|
+
``api_configuration`` carrying ``azimuth_labels`` (seed frame ->
|
|
52
|
+
label) and ``azimuth_label_meta`` (seed frame -> fit detail).
|
|
53
|
+
|
|
54
|
+
Returns:
|
|
55
|
+
DataFrame: The columns in :data:`RHEED_AZIMUTH_COLUMNS`, sorted by
|
|
56
|
+
``data_id`` then ``seed_frame``. ``azimuth_label`` is ``None`` for a seed
|
|
57
|
+
frame the classifier could not label, so callers can fall back to the
|
|
58
|
+
angle rather than inventing an identity. Non-rotating or unconfigured
|
|
59
|
+
entries contribute no rows; an empty payload yields an empty frame.
|
|
60
|
+
"""
|
|
61
|
+
rows: list[dict[str, Any]] = []
|
|
62
|
+
for entry in payload or []:
|
|
63
|
+
data_id = entry.get("data_id")
|
|
64
|
+
angles = entry.get("seed_frame_angles") or {}
|
|
65
|
+
api_configuration = entry.get("api_configuration") or {}
|
|
66
|
+
labels = api_configuration.get("azimuth_labels") or {}
|
|
67
|
+
meta = api_configuration.get("azimuth_label_meta") or {}
|
|
68
|
+
|
|
69
|
+
# Seed frames come from the angle map: it is the endpoint's own record of
|
|
70
|
+
# which frames this recording actually has, whereas the label map may be
|
|
71
|
+
# partial (an unlabelled azimuth is simply absent).
|
|
72
|
+
for seed_frame, angle in angles.items():
|
|
73
|
+
detail = meta.get(seed_frame) or {}
|
|
74
|
+
surface = detail.get("surface")
|
|
75
|
+
rows.append(
|
|
76
|
+
{
|
|
77
|
+
"data_id": str(data_id),
|
|
78
|
+
"seed_frame": int(seed_frame),
|
|
79
|
+
"angle_degrees": float(angle)
|
|
80
|
+
if angle is not None
|
|
81
|
+
else float("nan"),
|
|
82
|
+
"azimuth_label": labels.get(seed_frame),
|
|
83
|
+
"label_confidence": (
|
|
84
|
+
float(detail["confidence"])
|
|
85
|
+
if detail.get("confidence") is not None
|
|
86
|
+
else float("nan")
|
|
87
|
+
),
|
|
88
|
+
"crystal_system": detail.get("system"),
|
|
89
|
+
# A Miller index arrives as a list; join it so the column stays
|
|
90
|
+
# scalar and groupable.
|
|
91
|
+
"surface_miller": (
|
|
92
|
+
"".join(str(_i) for _i in surface)
|
|
93
|
+
if isinstance(surface, list | tuple)
|
|
94
|
+
else surface
|
|
95
|
+
),
|
|
96
|
+
}
|
|
97
|
+
)
|
|
98
|
+
|
|
99
|
+
if not rows:
|
|
100
|
+
return _empty()
|
|
101
|
+
|
|
102
|
+
return (
|
|
103
|
+
DataFrame(rows)[list(RHEED_AZIMUTH_COLUMNS)]
|
|
104
|
+
.sort_values(["data_id", "seed_frame"], kind="stable")
|
|
105
|
+
.reset_index(drop=True)
|
|
106
|
+
)
|
|
107
|
+
|
|
108
|
+
|
|
109
|
+
def azimuth_label_by_seed_frame(
|
|
110
|
+
azimuths: DataFrame, data_id: str
|
|
111
|
+
) -> dict[int, str | None]:
|
|
112
|
+
"""``{seed_frame: azimuth_label}`` for one data item, for joining onto series.
|
|
113
|
+
|
|
114
|
+
Args:
|
|
115
|
+
azimuths: A frame from :func:`rheed_azimuths_to_dataframe`.
|
|
116
|
+
data_id: The data item to select.
|
|
117
|
+
|
|
118
|
+
Returns:
|
|
119
|
+
dict: Seed frame -> label, including seed frames whose label is ``None``.
|
|
120
|
+
"""
|
|
121
|
+
if not len(azimuths):
|
|
122
|
+
return {}
|
|
123
|
+
mine = azimuths[azimuths["data_id"].astype(str) == str(data_id)]
|
|
124
|
+
return {
|
|
125
|
+
int(row.seed_frame): (None if pd.isna(row.azimuth_label) else row.azimuth_label)
|
|
126
|
+
for row in mine.itertuples()
|
|
127
|
+
}
|
|
@@ -0,0 +1,292 @@
|
|
|
1
|
+
"""Parse the sample-scoped payloads that are *not* timeseries.
|
|
2
|
+
|
|
3
|
+
Three read endpoints describe a physical sample without returning a series:
|
|
4
|
+
|
|
5
|
+
* ``GET /physical_samples/{id}/process_steps`` — the sample's temporal phases
|
|
6
|
+
(one growth/anneal/cooldown segment each), derived by clustering its data
|
|
7
|
+
items' acquisition windows.
|
|
8
|
+
* ``GET /physical_samples/{id}/annotations/`` — spatially-resolved property
|
|
9
|
+
values pinned to surface coordinates (e.g. an XPS ``sr_ti_ratio`` measured
|
|
10
|
+
every 22.5 mm across a wafer).
|
|
11
|
+
* ``GET /summary/{data_id}`` — the agent-written summary of one data item,
|
|
12
|
+
together with the status of the summarization workflow.
|
|
13
|
+
|
|
14
|
+
Each helper converts one decoded response body into a flat DataFrame. Sample
|
|
15
|
+
timeseries live in :mod:`atomscale.timeseries.physical_sample` instead; these
|
|
16
|
+
payloads are metadata and event records, so they share no axis and are kept
|
|
17
|
+
apart.
|
|
18
|
+
"""
|
|
19
|
+
|
|
20
|
+
from __future__ import annotations
|
|
21
|
+
|
|
22
|
+
from collections.abc import Mapping, Sequence
|
|
23
|
+
from typing import Any
|
|
24
|
+
|
|
25
|
+
import pandas as pd
|
|
26
|
+
from pandas import DataFrame
|
|
27
|
+
|
|
28
|
+
# Process steps: one row per step, ordered by ``step_index``.
|
|
29
|
+
#
|
|
30
|
+
# ``data_ids`` stays in-frame as a tuple rather than moving to ``DataFrame.attrs``
|
|
31
|
+
# (the convention :mod:`atomscale.timeseries.physical_sample` uses for its
|
|
32
|
+
# provenance ids) because step membership *is* the payload here — callers fetch
|
|
33
|
+
# steps precisely to learn which data items belong to which phase — and a step
|
|
34
|
+
# with no members has to survive as a row.
|
|
35
|
+
PROCESS_STEP_COLUMNS: tuple[str, ...] = (
|
|
36
|
+
"step_index",
|
|
37
|
+
"process_step_id",
|
|
38
|
+
"name",
|
|
39
|
+
"start_unix_ms_utc",
|
|
40
|
+
"end_unix_ms_utc",
|
|
41
|
+
"duration_seconds",
|
|
42
|
+
"start_datetime",
|
|
43
|
+
"end_datetime",
|
|
44
|
+
"n_data_ids",
|
|
45
|
+
"data_ids",
|
|
46
|
+
"last_updated",
|
|
47
|
+
)
|
|
48
|
+
|
|
49
|
+
# Spatial annotations: one row per annotated point.
|
|
50
|
+
SPATIAL_ANNOTATION_COLUMNS: tuple[str, ...] = (
|
|
51
|
+
"annotation_id",
|
|
52
|
+
"data_id",
|
|
53
|
+
"char_source_type",
|
|
54
|
+
"property_name",
|
|
55
|
+
"property_value",
|
|
56
|
+
"property_unit",
|
|
57
|
+
"coord_x",
|
|
58
|
+
"coord_y",
|
|
59
|
+
"coord_z",
|
|
60
|
+
"coord_ref_frame",
|
|
61
|
+
"coord_units",
|
|
62
|
+
"metadata",
|
|
63
|
+
"last_updated",
|
|
64
|
+
)
|
|
65
|
+
|
|
66
|
+
# AI summaries: one row per *requested* data id, so a data item with no summary
|
|
67
|
+
# is visible as a ``task_status`` row rather than a missing row.
|
|
68
|
+
AI_SUMMARY_COLUMNS: tuple[str, ...] = (
|
|
69
|
+
"data_id",
|
|
70
|
+
"task_status",
|
|
71
|
+
"summary_id",
|
|
72
|
+
"content",
|
|
73
|
+
"structured_content",
|
|
74
|
+
"generation_model",
|
|
75
|
+
"generation_timestamp",
|
|
76
|
+
"user_feedback",
|
|
77
|
+
"workflow_id",
|
|
78
|
+
)
|
|
79
|
+
|
|
80
|
+
_FLOAT_COLUMNS: Mapping[str, tuple[str, ...]] = {
|
|
81
|
+
"process_steps": (
|
|
82
|
+
"start_unix_ms_utc",
|
|
83
|
+
"end_unix_ms_utc",
|
|
84
|
+
"duration_seconds",
|
|
85
|
+
),
|
|
86
|
+
"annotations": ("property_value", "coord_x", "coord_y", "coord_z"),
|
|
87
|
+
}
|
|
88
|
+
|
|
89
|
+
# Absolute-timestamp columns per payload. Parsed with ``utc=True`` so offset-free
|
|
90
|
+
# values land tz-aware (UTC) rather than tz-naive, keeping them comparable with
|
|
91
|
+
# each other and with the UTC ``start_datetime`` / ``end_datetime`` bounds — a
|
|
92
|
+
# tz-naive column would raise on subtraction/comparison against a tz-aware one.
|
|
93
|
+
_DATETIME_COLUMNS: Mapping[str, tuple[str, ...]] = {
|
|
94
|
+
"process_steps": ("start_datetime", "end_datetime", "last_updated"),
|
|
95
|
+
"annotations": ("last_updated",),
|
|
96
|
+
"summaries": ("generation_timestamp",),
|
|
97
|
+
}
|
|
98
|
+
|
|
99
|
+
|
|
100
|
+
def _empty(
|
|
101
|
+
columns: Sequence[str],
|
|
102
|
+
float_columns: Sequence[str] = (),
|
|
103
|
+
datetime_columns: Sequence[str] = (),
|
|
104
|
+
) -> DataFrame:
|
|
105
|
+
"""Empty frame carrying the full column set, with typed float/datetime columns.
|
|
106
|
+
|
|
107
|
+
Datetime columns are typed tz-aware (UTC) so an empty return shares the dtype
|
|
108
|
+
the populated path produces (see :data:`_DATETIME_COLUMNS`).
|
|
109
|
+
"""
|
|
110
|
+
frame = DataFrame({c: [] for c in columns})
|
|
111
|
+
for column in float_columns:
|
|
112
|
+
frame[column] = frame[column].astype("float64")
|
|
113
|
+
for column in datetime_columns:
|
|
114
|
+
frame[column] = frame[column].astype("datetime64[ns, UTC]")
|
|
115
|
+
return frame
|
|
116
|
+
|
|
117
|
+
|
|
118
|
+
def _as_float(value: Any) -> float:
|
|
119
|
+
"""Coerce a JSON number (or ``None``) to ``float``, ``None`` -> ``NaN``."""
|
|
120
|
+
return float("nan") if value is None else float(value)
|
|
121
|
+
|
|
122
|
+
|
|
123
|
+
def process_steps_to_dataframe(payload: Mapping[str, Any]) -> DataFrame:
|
|
124
|
+
"""Convert a ``ProcessStepsResponse`` payload to one row per process step.
|
|
125
|
+
|
|
126
|
+
Args:
|
|
127
|
+
payload: Decoded response body, with a ``"process_steps"`` list whose
|
|
128
|
+
entries carry ``id``, ``step_index``, ``name``,
|
|
129
|
+
``start_unix_ms_utc`` / ``end_unix_ms_utc`` (float epoch
|
|
130
|
+
milliseconds, UTC), ``data_ids``, and ``last_updated``.
|
|
131
|
+
|
|
132
|
+
Returns:
|
|
133
|
+
DataFrame: The columns in :data:`PROCESS_STEP_COLUMNS`, sorted by
|
|
134
|
+
``step_index``. ``duration_seconds`` and the tz-aware
|
|
135
|
+
``start_datetime`` / ``end_datetime`` are derived from the epoch-ms
|
|
136
|
+
bounds for convenience; the raw bounds are kept so callers aligning
|
|
137
|
+
against absolute-time series (RHEED shutter times, tool state) never
|
|
138
|
+
have to round-trip through a datetime. A sample with no steps yields an
|
|
139
|
+
empty frame, never an error.
|
|
140
|
+
"""
|
|
141
|
+
steps = list(payload.get("process_steps") or [])
|
|
142
|
+
if not steps:
|
|
143
|
+
return _empty(
|
|
144
|
+
PROCESS_STEP_COLUMNS,
|
|
145
|
+
_FLOAT_COLUMNS["process_steps"],
|
|
146
|
+
_DATETIME_COLUMNS["process_steps"],
|
|
147
|
+
)
|
|
148
|
+
|
|
149
|
+
rows = []
|
|
150
|
+
for step in steps:
|
|
151
|
+
start = _as_float(step.get("start_unix_ms_utc"))
|
|
152
|
+
end = _as_float(step.get("end_unix_ms_utc"))
|
|
153
|
+
data_ids = tuple(step.get("data_ids") or ())
|
|
154
|
+
rows.append(
|
|
155
|
+
{
|
|
156
|
+
"step_index": step.get("step_index"),
|
|
157
|
+
"process_step_id": step.get("id"),
|
|
158
|
+
"name": step.get("name"),
|
|
159
|
+
"start_unix_ms_utc": start,
|
|
160
|
+
"end_unix_ms_utc": end,
|
|
161
|
+
"duration_seconds": (end - start) / 1000.0,
|
|
162
|
+
"n_data_ids": len(data_ids),
|
|
163
|
+
"data_ids": data_ids,
|
|
164
|
+
"last_updated": step.get("last_updated"),
|
|
165
|
+
}
|
|
166
|
+
)
|
|
167
|
+
|
|
168
|
+
frame = DataFrame(rows)
|
|
169
|
+
frame["start_datetime"] = pd.to_datetime(
|
|
170
|
+
frame["start_unix_ms_utc"], unit="ms", utc=True
|
|
171
|
+
)
|
|
172
|
+
frame["end_datetime"] = pd.to_datetime(
|
|
173
|
+
frame["end_unix_ms_utc"], unit="ms", utc=True
|
|
174
|
+
)
|
|
175
|
+
frame["last_updated"] = pd.to_datetime(
|
|
176
|
+
frame["last_updated"], errors="coerce", utc=True
|
|
177
|
+
)
|
|
178
|
+
return (
|
|
179
|
+
frame[list(PROCESS_STEP_COLUMNS)]
|
|
180
|
+
.sort_values("step_index", kind="stable")
|
|
181
|
+
.reset_index(drop=True)
|
|
182
|
+
)
|
|
183
|
+
|
|
184
|
+
|
|
185
|
+
def spatial_annotations_to_dataframe(payload: Sequence[Mapping[str, Any]]) -> DataFrame:
|
|
186
|
+
"""Convert a ``list[SpatialAnnotationRead]`` payload to one row per point.
|
|
187
|
+
|
|
188
|
+
Args:
|
|
189
|
+
payload: Decoded response body — a list of annotation objects, each with
|
|
190
|
+
``id``, ``property_name``, ``property_value``, optional
|
|
191
|
+
``property_unit`` / ``char_source_type`` / ``data_id``, the
|
|
192
|
+
``coord_x`` / ``coord_y`` / ``coord_z`` surface position with its
|
|
193
|
+
``coord_ref_frame`` and ``coord_units``, and ``metadata``.
|
|
194
|
+
|
|
195
|
+
Returns:
|
|
196
|
+
DataFrame: The columns in :data:`SPATIAL_ANNOTATION_COLUMNS`, sorted by
|
|
197
|
+
``property_name`` then position so a scan reads left-to-right. The
|
|
198
|
+
server-side ``id`` is exposed as ``annotation_id`` to keep it distinct
|
|
199
|
+
from ``data_id``. No annotations yields an empty frame, never an error.
|
|
200
|
+
"""
|
|
201
|
+
records = list(payload or [])
|
|
202
|
+
if not records:
|
|
203
|
+
return _empty(
|
|
204
|
+
SPATIAL_ANNOTATION_COLUMNS,
|
|
205
|
+
_FLOAT_COLUMNS["annotations"],
|
|
206
|
+
_DATETIME_COLUMNS["annotations"],
|
|
207
|
+
)
|
|
208
|
+
|
|
209
|
+
rows = [
|
|
210
|
+
{
|
|
211
|
+
"annotation_id": record.get("id"),
|
|
212
|
+
"data_id": record.get("data_id"),
|
|
213
|
+
"char_source_type": record.get("char_source_type"),
|
|
214
|
+
"property_name": record.get("property_name"),
|
|
215
|
+
"property_value": _as_float(record.get("property_value")),
|
|
216
|
+
"property_unit": record.get("property_unit"),
|
|
217
|
+
"coord_x": _as_float(record.get("coord_x")),
|
|
218
|
+
"coord_y": _as_float(record.get("coord_y")),
|
|
219
|
+
"coord_z": _as_float(record.get("coord_z")),
|
|
220
|
+
"coord_ref_frame": record.get("coord_ref_frame"),
|
|
221
|
+
"coord_units": record.get("coord_units"),
|
|
222
|
+
"metadata": record.get("metadata"),
|
|
223
|
+
"last_updated": record.get("last_updated"),
|
|
224
|
+
}
|
|
225
|
+
for record in records
|
|
226
|
+
]
|
|
227
|
+
|
|
228
|
+
frame = DataFrame(rows)
|
|
229
|
+
frame["last_updated"] = pd.to_datetime(
|
|
230
|
+
frame["last_updated"], errors="coerce", utc=True
|
|
231
|
+
)
|
|
232
|
+
return (
|
|
233
|
+
frame[list(SPATIAL_ANNOTATION_COLUMNS)]
|
|
234
|
+
.sort_values(
|
|
235
|
+
["property_name", "coord_x", "coord_y"], kind="stable", na_position="last"
|
|
236
|
+
)
|
|
237
|
+
.reset_index(drop=True)
|
|
238
|
+
)
|
|
239
|
+
|
|
240
|
+
|
|
241
|
+
def ai_summary_row(data_id: str, payload: Mapping[str, Any] | None) -> dict[str, Any]:
|
|
242
|
+
"""Flatten one ``GetSummaryResponse`` into an :data:`AI_SUMMARY_COLUMNS` row.
|
|
243
|
+
|
|
244
|
+
Args:
|
|
245
|
+
data_id: The data id the response was fetched for; carried through so a
|
|
246
|
+
batch of rows stays attributable.
|
|
247
|
+
payload: Decoded response body with ``summary`` (the note, or ``None``),
|
|
248
|
+
``task_status`` (``"available"`` / ``"pending"`` / ``"not_found"``),
|
|
249
|
+
and ``workflow_id``. ``None`` — a 404 from the endpoint — is
|
|
250
|
+
reported as ``task_status="not_found"``, matching how the endpoint
|
|
251
|
+
reports an un-summarized data item that does exist.
|
|
252
|
+
|
|
253
|
+
Returns:
|
|
254
|
+
dict: One flat row; summary-note fields are ``None`` unless the note is
|
|
255
|
+
present.
|
|
256
|
+
"""
|
|
257
|
+
body = payload or {}
|
|
258
|
+
summary = body.get("summary") or {}
|
|
259
|
+
return {
|
|
260
|
+
"data_id": data_id,
|
|
261
|
+
"task_status": body.get("task_status") or "not_found",
|
|
262
|
+
"summary_id": summary.get("id"),
|
|
263
|
+
"content": summary.get("content"),
|
|
264
|
+
"structured_content": summary.get("structured_content"),
|
|
265
|
+
"generation_model": summary.get("generation_model"),
|
|
266
|
+
"generation_timestamp": summary.get("generation_timestamp"),
|
|
267
|
+
"user_feedback": summary.get("user_feedback"),
|
|
268
|
+
"workflow_id": body.get("workflow_id"),
|
|
269
|
+
}
|
|
270
|
+
|
|
271
|
+
|
|
272
|
+
def ai_summaries_to_dataframe(rows: Sequence[Mapping[str, Any]]) -> DataFrame:
|
|
273
|
+
"""Assemble :func:`ai_summary_row` outputs into a frame.
|
|
274
|
+
|
|
275
|
+
Args:
|
|
276
|
+
rows: One row per requested data id, in the caller's request order.
|
|
277
|
+
|
|
278
|
+
Returns:
|
|
279
|
+
DataFrame: The columns in :data:`AI_SUMMARY_COLUMNS`, request order
|
|
280
|
+
preserved, with ``generation_timestamp`` parsed to datetime. An empty
|
|
281
|
+
``rows`` yields an empty frame.
|
|
282
|
+
"""
|
|
283
|
+
if not rows:
|
|
284
|
+
return _empty(
|
|
285
|
+
AI_SUMMARY_COLUMNS, datetime_columns=_DATETIME_COLUMNS["summaries"]
|
|
286
|
+
)
|
|
287
|
+
|
|
288
|
+
frame = DataFrame(list(rows))[list(AI_SUMMARY_COLUMNS)]
|
|
289
|
+
frame["generation_timestamp"] = pd.to_datetime(
|
|
290
|
+
frame["generation_timestamp"], errors="coerce", utc=True
|
|
291
|
+
)
|
|
292
|
+
return frame
|
|
@@ -61,6 +61,8 @@ scripts/ci_auth_preflight.py
|
|
|
61
61
|
src/atomicds/__init__.py
|
|
62
62
|
src/atomscale/__init__.py
|
|
63
63
|
src/atomscale/client.py
|
|
64
|
+
src/atomscale/rheed_metadata.py
|
|
65
|
+
src/atomscale/samples.py
|
|
64
66
|
src/atomscale.egg-info/PKG-INFO
|
|
65
67
|
src/atomscale.egg-info/SOURCES.txt
|
|
66
68
|
src/atomscale.egg-info/dependency_links.txt
|
|
@@ -139,6 +141,7 @@ tests/test_rheed_image.py
|
|
|
139
141
|
tests/test_rheed_stream.py
|
|
140
142
|
tests/test_rheed_timeseries.py
|
|
141
143
|
tests/test_rheed_video.py
|
|
144
|
+
tests/test_samples.py
|
|
142
145
|
tests/test_similarity_matches.py
|
|
143
146
|
tests/test_similarity_trajectory.py
|
|
144
147
|
tests/test_timeseries_stream.py
|
|
@@ -83,6 +83,8 @@
|
|
|
83
83
|
"src/atomscale/results/unknown.py",
|
|
84
84
|
"src/atomscale/results/xps.py",
|
|
85
85
|
"src/atomscale/results/xrd.py",
|
|
86
|
+
"src/atomscale/rheed_metadata.py",
|
|
87
|
+
"src/atomscale/samples.py",
|
|
86
88
|
"src/atomscale/similarity/__init__.py",
|
|
87
89
|
"src/atomscale/similarity/embedding_provider.py",
|
|
88
90
|
"src/atomscale/similarity/polling.py",
|
|
@@ -134,6 +136,7 @@
|
|
|
134
136
|
"tests/test_rheed_stream.py",
|
|
135
137
|
"tests/test_rheed_timeseries.py",
|
|
136
138
|
"tests/test_rheed_video.py",
|
|
139
|
+
"tests/test_samples.py",
|
|
137
140
|
"tests/test_similarity_matches.py",
|
|
138
141
|
"tests/test_similarity_trajectory.py",
|
|
139
142
|
"tests/test_timeseries_stream.py",
|