Coverage for python/lsst/dax/apdb/cassandra/apdbCassandra.py: 18%
599 statements
« prev ^ index » next coverage.py v7.5.1, created at 2024-05-11 03:30 -0700
« prev ^ index » next coverage.py v7.5.1, created at 2024-05-11 03:30 -0700
1# This file is part of dax_apdb.
2#
3# Developed for the LSST Data Management System.
4# This product includes software developed by the LSST Project
5# (http://www.lsst.org).
6# See the COPYRIGHT file at the top-level directory of this distribution
7# for details of code ownership.
8#
9# This program is free software: you can redistribute it and/or modify
10# it under the terms of the GNU General Public License as published by
11# the Free Software Foundation, either version 3 of the License, or
12# (at your option) any later version.
13#
14# This program is distributed in the hope that it will be useful,
15# but WITHOUT ANY WARRANTY; without even the implied warranty of
16# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
17# GNU General Public License for more details.
18#
19# You should have received a copy of the GNU General Public License
20# along with this program. If not, see <http://www.gnu.org/licenses/>.
22from __future__ import annotations
24__all__ = ["ApdbCassandraConfig", "ApdbCassandra"]
26import dataclasses
27import json
28import logging
29from collections.abc import Iterable, Iterator, Mapping, Set
30from typing import TYPE_CHECKING, Any, cast
32import numpy as np
33import pandas
35# If cassandra-driver is not there the module can still be imported
36# but ApdbCassandra cannot be instantiated.
37try:
38 import cassandra
39 import cassandra.query
40 from cassandra.auth import AuthProvider, PlainTextAuthProvider
41 from cassandra.cluster import EXEC_PROFILE_DEFAULT, Cluster, ExecutionProfile, Session
42 from cassandra.policies import AddressTranslator, RoundRobinPolicy, WhiteListRoundRobinPolicy
44 CASSANDRA_IMPORTED = True
45except ImportError:
46 CASSANDRA_IMPORTED = False
48import astropy.time
49import felis.datamodel
50from lsst import sphgeom
51from lsst.pex.config import ChoiceField, Field, ListField
52from lsst.utils.db_auth import DbAuth, DbAuthNotFoundError
53from lsst.utils.iteration import chunk_iterable
55from ..apdb import Apdb, ApdbConfig
56from ..apdbConfigFreezer import ApdbConfigFreezer
57from ..apdbReplica import ReplicaChunk
58from ..apdbSchema import ApdbTables
59from ..monitor import MonAgent
60from ..pixelization import Pixelization
61from ..schema_model import Table
62from ..timer import Timer
63from ..versionTuple import IncompatibleVersionError, VersionTuple
64from .apdbCassandraReplica import ApdbCassandraReplica
65from .apdbCassandraSchema import ApdbCassandraSchema, ExtraTables
66from .apdbMetadataCassandra import ApdbMetadataCassandra
67from .cassandra_utils import (
68 PreparedStatementCache,
69 literal,
70 pandas_dataframe_factory,
71 quote_id,
72 raw_data_factory,
73 select_concurrent,
74)
76if TYPE_CHECKING:
77 from ..apdbMetadata import ApdbMetadata
79_LOG = logging.getLogger(__name__)
81_MON = MonAgent(__name__)
83VERSION = VersionTuple(0, 1, 0)
84"""Version for the code controlling non-replication tables. This needs to be
85updated following compatibility rules when schema produced by this code
86changes.
87"""
89# Copied from daf_butler.
90DB_AUTH_ENVVAR = "LSST_DB_AUTH"
91"""Default name of the environmental variable that will be used to locate DB
92credentials configuration file. """
94DB_AUTH_PATH = "~/.lsst/db-auth.yaml"
95"""Default path at which it is expected that DB credentials are found."""
98class CassandraMissingError(Exception):
99 def __init__(self) -> None:
100 super().__init__("cassandra-driver module cannot be imported")
103class ApdbCassandraConfig(ApdbConfig):
104 """Configuration class for Cassandra-based APDB implementation."""
106 contact_points = ListField[str](
107 doc="The list of contact points to try connecting for cluster discovery.", default=["127.0.0.1"]
108 )
109 private_ips = ListField[str](doc="List of internal IP addresses for contact_points.", default=[])
110 port = Field[int](doc="Port number to connect to.", default=9042)
111 keyspace = Field[str](doc="Default keyspace for operations.", default="apdb")
112 username = Field[str](
113 doc=f"Cassandra user name, if empty then {DB_AUTH_PATH} has to provide it with password.",
114 default="",
115 )
116 read_consistency = Field[str](
117 doc="Name for consistency level of read operations, default: QUORUM, can be ONE.", default="QUORUM"
118 )
119 write_consistency = Field[str](
120 doc="Name for consistency level of write operations, default: QUORUM, can be ONE.", default="QUORUM"
121 )
122 read_timeout = Field[float](doc="Timeout in seconds for read operations.", default=120.0)
123 write_timeout = Field[float](doc="Timeout in seconds for write operations.", default=60.0)
124 remove_timeout = Field[float](doc="Timeout in seconds for remove operations.", default=600.0)
125 read_concurrency = Field[int](doc="Concurrency level for read operations.", default=500)
126 protocol_version = Field[int](
127 doc="Cassandra protocol version to use, default is V4",
128 default=cassandra.ProtocolVersion.V4 if CASSANDRA_IMPORTED else 0,
129 )
130 dia_object_columns = ListField[str](
131 doc="List of columns to read from DiaObject[Last], by default read all columns", default=[]
132 )
133 prefix = Field[str](doc="Prefix to add to table names", default="")
134 part_pixelization = ChoiceField[str](
135 allowed=dict(htm="HTM pixelization", q3c="Q3C pixelization", mq3c="MQ3C pixelization"),
136 doc="Pixelization used for partitioning index.",
137 default="mq3c",
138 )
139 part_pix_level = Field[int](doc="Pixelization level used for partitioning index.", default=10)
140 part_pix_max_ranges = Field[int](doc="Max number of ranges in pixelization envelope", default=64)
141 ra_dec_columns = ListField[str](default=["ra", "dec"], doc="Names of ra/dec columns in DiaObject table")
142 timer = Field[bool](doc="If True then print/log timing information", default=False)
143 time_partition_tables = Field[bool](
144 doc="Use per-partition tables for sources instead of partitioning by time", default=False
145 )
146 time_partition_days = Field[int](
147 doc=(
148 "Time partitioning granularity in days, this value must not be changed after database is "
149 "initialized"
150 ),
151 default=30,
152 )
153 time_partition_start = Field[str](
154 doc=(
155 "Starting time for per-partition tables, in yyyy-mm-ddThh:mm:ss format, in TAI. "
156 "This is used only when time_partition_tables is True."
157 ),
158 default="2018-12-01T00:00:00",
159 )
160 time_partition_end = Field[str](
161 doc=(
162 "Ending time for per-partition tables, in yyyy-mm-ddThh:mm:ss format, in TAI. "
163 "This is used only when time_partition_tables is True."
164 ),
165 default="2030-01-01T00:00:00",
166 )
167 query_per_time_part = Field[bool](
168 default=False,
169 doc=(
170 "If True then build separate query for each time partition, otherwise build one single query. "
171 "This is only used when time_partition_tables is False in schema config."
172 ),
173 )
174 query_per_spatial_part = Field[bool](
175 default=False,
176 doc="If True then build one query per spatial partition, otherwise build single query.",
177 )
178 use_insert_id_skips_diaobjects = Field[bool](
179 default=False,
180 doc=(
181 "If True then do not store DiaObjects when use_insert_id is True "
182 "(DiaObjectsChunks has the same data)."
183 ),
184 )
187@dataclasses.dataclass
188class _FrozenApdbCassandraConfig:
189 """Part of the configuration that is saved in metadata table and read back.
191 The attributes are a subset of attributes in `ApdbCassandraConfig` class.
193 Parameters
194 ----------
195 config : `ApdbSqlConfig`
196 Configuration used to copy initial values of attributes.
197 """
199 use_insert_id: bool
200 part_pixelization: str
201 part_pix_level: int
202 ra_dec_columns: list[str]
203 time_partition_tables: bool
204 time_partition_days: int
205 use_insert_id_skips_diaobjects: bool
207 def __init__(self, config: ApdbCassandraConfig):
208 self.use_insert_id = config.use_insert_id
209 self.part_pixelization = config.part_pixelization
210 self.part_pix_level = config.part_pix_level
211 self.ra_dec_columns = list(config.ra_dec_columns)
212 self.time_partition_tables = config.time_partition_tables
213 self.time_partition_days = config.time_partition_days
214 self.use_insert_id_skips_diaobjects = config.use_insert_id_skips_diaobjects
216 def to_json(self) -> str:
217 """Convert this instance to JSON representation."""
218 return json.dumps(dataclasses.asdict(self))
220 def update(self, json_str: str) -> None:
221 """Update attribute values from a JSON string.
223 Parameters
224 ----------
225 json_str : str
226 String containing JSON representation of configuration.
227 """
228 data = json.loads(json_str)
229 if not isinstance(data, dict):
230 raise TypeError(f"JSON string must be convertible to object: {json_str!r}")
231 allowed_names = {field.name for field in dataclasses.fields(self)}
232 for key, value in data.items():
233 if key not in allowed_names:
234 raise ValueError(f"JSON object contains unknown key: {key}")
235 setattr(self, key, value)
238if CASSANDRA_IMPORTED: 238 ↛ 240line 238 didn't jump to line 240, because the condition on line 238 was never true
240 class _AddressTranslator(AddressTranslator):
241 """Translate internal IP address to external.
243 Only used for docker-based setup, not viable long-term solution.
244 """
246 def __init__(self, public_ips: list[str], private_ips: list[str]):
247 self._map = dict((k, v) for k, v in zip(private_ips, public_ips))
249 def translate(self, private_ip: str) -> str:
250 return self._map.get(private_ip, private_ip)
253class ApdbCassandra(Apdb):
254 """Implementation of APDB database on to of Apache Cassandra.
256 The implementation is configured via standard ``pex_config`` mechanism
257 using `ApdbCassandraConfig` configuration class. For an example of
258 different configurations check config/ folder.
260 Parameters
261 ----------
262 config : `ApdbCassandraConfig`
263 Configuration object.
264 """
266 metadataSchemaVersionKey = "version:schema"
267 """Name of the metadata key to store schema version number."""
269 metadataCodeVersionKey = "version:ApdbCassandra"
270 """Name of the metadata key to store code version number."""
272 metadataReplicaVersionKey = "version:ApdbCassandraReplica"
273 """Name of the metadata key to store replica code version number."""
275 metadataConfigKey = "config:apdb-cassandra.json"
276 """Name of the metadata key to store code version number."""
278 _frozen_parameters = (
279 "use_insert_id",
280 "part_pixelization",
281 "part_pix_level",
282 "ra_dec_columns",
283 "time_partition_tables",
284 "time_partition_days",
285 "use_insert_id_skips_diaobjects",
286 )
287 """Names of the config parameters to be frozen in metadata table."""
289 partition_zero_epoch = astropy.time.Time(0, format="unix_tai")
290 """Start time for partition 0, this should never be changed."""
292 def __init__(self, config: ApdbCassandraConfig):
293 if not CASSANDRA_IMPORTED:
294 raise CassandraMissingError()
296 self._keyspace = config.keyspace
298 self._cluster, self._session = self._make_session(config)
300 meta_table_name = ApdbTables.metadata.table_name(config.prefix)
301 self._metadata = ApdbMetadataCassandra(
302 self._session, meta_table_name, config.keyspace, "read_tuples", "write"
303 )
305 # Read frozen config from metadata.
306 config_json = self._metadata.get(self.metadataConfigKey)
307 if config_json is not None:
308 # Update config from metadata.
309 freezer = ApdbConfigFreezer[ApdbCassandraConfig](self._frozen_parameters)
310 self.config = freezer.update(config, config_json)
311 else:
312 self.config = config
313 self.config.validate()
315 self._pixelization = Pixelization(
316 self.config.part_pixelization,
317 self.config.part_pix_level,
318 config.part_pix_max_ranges,
319 )
321 self._schema = ApdbCassandraSchema(
322 session=self._session,
323 keyspace=self._keyspace,
324 schema_file=self.config.schema_file,
325 schema_name=self.config.schema_name,
326 prefix=self.config.prefix,
327 time_partition_tables=self.config.time_partition_tables,
328 enable_replica=self.config.use_insert_id,
329 )
330 self._partition_zero_epoch_mjd = float(self.partition_zero_epoch.mjd)
332 if self._metadata.table_exists():
333 self._versionCheck(self._metadata)
335 # Cache for prepared statements
336 self._preparer = PreparedStatementCache(self._session)
338 _LOG.debug("ApdbCassandra Configuration:")
339 for key, value in self.config.items():
340 _LOG.debug(" %s: %s", key, value)
342 self._timer_args: list[MonAgent | logging.Logger] = [_MON]
343 if self.config.timer:
344 self._timer_args.append(_LOG)
346 def __del__(self) -> None:
347 if hasattr(self, "_cluster"):
348 self._cluster.shutdown()
350 def _timer(self, name: str, *, tags: Mapping[str, str | int] | None = None) -> Timer:
351 """Create `Timer` instance given its name."""
352 return Timer(name, *self._timer_args, tags=tags)
354 @classmethod
355 def _make_session(cls, config: ApdbCassandraConfig) -> tuple[Cluster, Session]:
356 """Make Cassandra session."""
357 addressTranslator: AddressTranslator | None = None
358 if config.private_ips:
359 addressTranslator = _AddressTranslator(list(config.contact_points), list(config.private_ips))
361 cluster = Cluster(
362 execution_profiles=cls._makeProfiles(config),
363 contact_points=config.contact_points,
364 port=config.port,
365 address_translator=addressTranslator,
366 protocol_version=config.protocol_version,
367 auth_provider=cls._make_auth_provider(config),
368 )
369 session = cluster.connect()
370 # Disable result paging
371 session.default_fetch_size = None
373 return cluster, session
375 @classmethod
376 def _make_auth_provider(cls, config: ApdbCassandraConfig) -> AuthProvider | None:
377 """Make Cassandra authentication provider instance."""
378 try:
379 dbauth = DbAuth(DB_AUTH_PATH, DB_AUTH_ENVVAR)
380 except DbAuthNotFoundError:
381 # Credentials file doesn't exist, use anonymous login.
382 return None
384 empty_username = True
385 # Try every contact point in turn.
386 for hostname in config.contact_points:
387 try:
388 username, password = dbauth.getAuth(
389 "cassandra", config.username, hostname, config.port, config.keyspace
390 )
391 if not username:
392 # Password without user name, try next hostname, but give
393 # warning later if no better match is found.
394 empty_username = True
395 else:
396 return PlainTextAuthProvider(username=username, password=password)
397 except DbAuthNotFoundError:
398 pass
400 if empty_username:
401 _LOG.warning(
402 f"Credentials file ({DB_AUTH_PATH} or ${DB_AUTH_ENVVAR}) provided password but not "
403 f"user name, anonymous Cassandra logon will be attempted."
404 )
406 return None
408 def _versionCheck(self, metadata: ApdbMetadataCassandra) -> None:
409 """Check schema version compatibility."""
411 def _get_version(key: str, default: VersionTuple) -> VersionTuple:
412 """Retrieve version number from given metadata key."""
413 if metadata.table_exists():
414 version_str = metadata.get(key)
415 if version_str is None:
416 # Should not happen with existing metadata table.
417 raise RuntimeError(f"Version key {key!r} does not exist in metadata table.")
418 return VersionTuple.fromString(version_str)
419 return default
421 # For old databases where metadata table does not exist we assume that
422 # version of both code and schema is 0.1.0.
423 initial_version = VersionTuple(0, 1, 0)
424 db_schema_version = _get_version(self.metadataSchemaVersionKey, initial_version)
425 db_code_version = _get_version(self.metadataCodeVersionKey, initial_version)
427 # For now there is no way to make read-only APDB instances, assume that
428 # any access can do updates.
429 if not self._schema.schemaVersion().checkCompatibility(db_schema_version):
430 raise IncompatibleVersionError(
431 f"Configured schema version {self._schema.schemaVersion()} "
432 f"is not compatible with database version {db_schema_version}"
433 )
434 if not self.apdbImplementationVersion().checkCompatibility(db_code_version):
435 raise IncompatibleVersionError(
436 f"Current code version {self.apdbImplementationVersion()} "
437 f"is not compatible with database version {db_code_version}"
438 )
440 # Check replica code version only if replica is enabled.
441 if self._schema.has_replica_chunks:
442 db_replica_version = _get_version(self.metadataReplicaVersionKey, initial_version)
443 code_replica_version = ApdbCassandraReplica.apdbReplicaImplementationVersion()
444 if not code_replica_version.checkCompatibility(db_replica_version):
445 raise IncompatibleVersionError(
446 f"Current replication code version {code_replica_version} "
447 f"is not compatible with database version {db_replica_version}"
448 )
450 @classmethod
451 def apdbImplementationVersion(cls) -> VersionTuple:
452 """Return version number for current APDB implementation.
454 Returns
455 -------
456 version : `VersionTuple`
457 Version of the code defined in implementation class.
458 """
459 return VERSION
461 def tableDef(self, table: ApdbTables) -> Table | None:
462 # docstring is inherited from a base class
463 return self._schema.tableSchemas.get(table)
465 @classmethod
466 def init_database(
467 cls,
468 hosts: list[str],
469 keyspace: str,
470 *,
471 schema_file: str | None = None,
472 schema_name: str | None = None,
473 read_sources_months: int | None = None,
474 read_forced_sources_months: int | None = None,
475 use_insert_id: bool = False,
476 use_insert_id_skips_diaobjects: bool = False,
477 port: int | None = None,
478 username: str | None = None,
479 prefix: str | None = None,
480 part_pixelization: str | None = None,
481 part_pix_level: int | None = None,
482 time_partition_tables: bool = True,
483 time_partition_start: str | None = None,
484 time_partition_end: str | None = None,
485 read_consistency: str | None = None,
486 write_consistency: str | None = None,
487 read_timeout: int | None = None,
488 write_timeout: int | None = None,
489 ra_dec_columns: list[str] | None = None,
490 replication_factor: int | None = None,
491 drop: bool = False,
492 ) -> ApdbCassandraConfig:
493 """Initialize new APDB instance and make configuration object for it.
495 Parameters
496 ----------
497 hosts : `list` [`str`]
498 List of host names or IP addresses for Cassandra cluster.
499 keyspace : `str`
500 Name of the keyspace for APDB tables.
501 schema_file : `str`, optional
502 Location of (YAML) configuration file with APDB schema. If not
503 specified then default location will be used.
504 schema_name : `str`, optional
505 Name of the schema in YAML configuration file. If not specified
506 then default name will be used.
507 read_sources_months : `int`, optional
508 Number of months of history to read from DiaSource.
509 read_forced_sources_months : `int`, optional
510 Number of months of history to read from DiaForcedSource.
511 use_insert_id : `bool`, optional
512 If True, make additional tables used for replication to PPDB.
513 use_insert_id_skips_diaobjects : `bool`, optional
514 If `True` then do not fill regular ``DiaObject`` table when
515 ``use_insert_id`` is `True`.
516 port : `int`, optional
517 Port number to use for Cassandra connections.
518 username : `str`, optional
519 User name for Cassandra connections.
520 prefix : `str`, optional
521 Optional prefix for all table names.
522 part_pixelization : `str`, optional
523 Name of the MOC pixelization used for partitioning.
524 part_pix_level : `int`, optional
525 Pixelization level.
526 time_partition_tables : `bool`, optional
527 Create per-partition tables.
528 time_partition_start : `str`, optional
529 Starting time for per-partition tables, in yyyy-mm-ddThh:mm:ss
530 format, in TAI.
531 time_partition_end : `str`, optional
532 Ending time for per-partition tables, in yyyy-mm-ddThh:mm:ss
533 format, in TAI.
534 read_consistency : `str`, optional
535 Name of the consistency level for read operations.
536 write_consistency : `str`, optional
537 Name of the consistency level for write operations.
538 read_timeout : `int`, optional
539 Read timeout in seconds.
540 write_timeout : `int`, optional
541 Write timeout in seconds.
542 ra_dec_columns : `list` [`str`], optional
543 Names of ra/dec columns in DiaObject table.
544 replication_factor : `int`, optional
545 Replication factor used when creating new keyspace, if keyspace
546 already exists its replication factor is not changed.
547 drop : `bool`, optional
548 If `True` then drop existing tables before re-creating the schema.
550 Returns
551 -------
552 config : `ApdbCassandraConfig`
553 Resulting configuration object for a created APDB instance.
554 """
555 config = ApdbCassandraConfig(
556 contact_points=hosts,
557 keyspace=keyspace,
558 use_insert_id=use_insert_id,
559 use_insert_id_skips_diaobjects=use_insert_id_skips_diaobjects,
560 time_partition_tables=time_partition_tables,
561 )
562 if schema_file is not None:
563 config.schema_file = schema_file
564 if schema_name is not None:
565 config.schema_name = schema_name
566 if read_sources_months is not None:
567 config.read_sources_months = read_sources_months
568 if read_forced_sources_months is not None:
569 config.read_forced_sources_months = read_forced_sources_months
570 if port is not None:
571 config.port = port
572 if username is not None:
573 config.username = username
574 if prefix is not None:
575 config.prefix = prefix
576 if part_pixelization is not None:
577 config.part_pixelization = part_pixelization
578 if part_pix_level is not None:
579 config.part_pix_level = part_pix_level
580 if time_partition_start is not None:
581 config.time_partition_start = time_partition_start
582 if time_partition_end is not None:
583 config.time_partition_end = time_partition_end
584 if read_consistency is not None:
585 config.read_consistency = read_consistency
586 if write_consistency is not None:
587 config.write_consistency = write_consistency
588 if read_timeout is not None:
589 config.read_timeout = read_timeout
590 if write_timeout is not None:
591 config.write_timeout = write_timeout
592 if ra_dec_columns is not None:
593 config.ra_dec_columns = ra_dec_columns
595 cls._makeSchema(config, drop=drop, replication_factor=replication_factor)
597 return config
599 def get_replica(self) -> ApdbCassandraReplica:
600 """Return `ApdbReplica` instance for this database."""
601 # Note that this instance has to stay alive while replica exists, so
602 # we pass reference to self.
603 return ApdbCassandraReplica(self, self._schema, self._session)
605 @classmethod
606 def _makeSchema(
607 cls, config: ApdbConfig, *, drop: bool = False, replication_factor: int | None = None
608 ) -> None:
609 # docstring is inherited from a base class
611 if not isinstance(config, ApdbCassandraConfig):
612 raise TypeError(f"Unexpected type of configuration object: {type(config)}")
614 cluster, session = cls._make_session(config)
616 schema = ApdbCassandraSchema(
617 session=session,
618 keyspace=config.keyspace,
619 schema_file=config.schema_file,
620 schema_name=config.schema_name,
621 prefix=config.prefix,
622 time_partition_tables=config.time_partition_tables,
623 enable_replica=config.use_insert_id,
624 )
626 # Ask schema to create all tables.
627 if config.time_partition_tables:
628 time_partition_start = astropy.time.Time(config.time_partition_start, format="isot", scale="tai")
629 time_partition_end = astropy.time.Time(config.time_partition_end, format="isot", scale="tai")
630 part_epoch = float(cls.partition_zero_epoch.mjd)
631 part_days = config.time_partition_days
632 part_range = (
633 cls._time_partition_cls(time_partition_start, part_epoch, part_days),
634 cls._time_partition_cls(time_partition_end, part_epoch, part_days) + 1,
635 )
636 schema.makeSchema(drop=drop, part_range=part_range, replication_factor=replication_factor)
637 else:
638 schema.makeSchema(drop=drop, replication_factor=replication_factor)
640 meta_table_name = ApdbTables.metadata.table_name(config.prefix)
641 metadata = ApdbMetadataCassandra(session, meta_table_name, config.keyspace, "read_tuples", "write")
643 # Fill version numbers, overrides if they existed before.
644 if metadata.table_exists():
645 metadata.set(cls.metadataSchemaVersionKey, str(schema.schemaVersion()), force=True)
646 metadata.set(cls.metadataCodeVersionKey, str(cls.apdbImplementationVersion()), force=True)
648 if config.use_insert_id:
649 # Only store replica code version if replica is enabled.
650 metadata.set(
651 cls.metadataReplicaVersionKey,
652 str(ApdbCassandraReplica.apdbReplicaImplementationVersion()),
653 force=True,
654 )
656 # Store frozen part of a configuration in metadata.
657 freezer = ApdbConfigFreezer[ApdbCassandraConfig](cls._frozen_parameters)
658 metadata.set(cls.metadataConfigKey, freezer.to_json(config), force=True)
660 cluster.shutdown()
662 def getDiaObjects(self, region: sphgeom.Region) -> pandas.DataFrame:
663 # docstring is inherited from a base class
665 sp_where = self._spatial_where(region)
666 _LOG.debug("getDiaObjects: #partitions: %s", len(sp_where))
668 # We need to exclude extra partitioning columns from result.
669 column_names = self._schema.apdbColumnNames(ApdbTables.DiaObjectLast)
670 what = ",".join(quote_id(column) for column in column_names)
672 table_name = self._schema.tableName(ApdbTables.DiaObjectLast)
673 query = f'SELECT {what} from "{self._keyspace}"."{table_name}"'
674 statements: list[tuple] = []
675 for where, params in sp_where:
676 full_query = f"{query} WHERE {where}"
677 if params:
678 statement = self._preparer.prepare(full_query)
679 else:
680 # If there are no params then it is likely that query has a
681 # bunch of literals rendered already, no point trying to
682 # prepare it because it's not reusable.
683 statement = cassandra.query.SimpleStatement(full_query)
684 statements.append((statement, params))
685 _LOG.debug("getDiaObjects: #queries: %s", len(statements))
687 with _MON.context_tags({"table": "DiaObject"}):
688 _MON.add_record(
689 "select_query_stats", values={"num_sp_part": len(sp_where), "num_queries": len(statements)}
690 )
691 with self._timer("select_time"):
692 objects = cast(
693 pandas.DataFrame,
694 select_concurrent(
695 self._session, statements, "read_pandas_multi", self.config.read_concurrency
696 ),
697 )
699 _LOG.debug("found %s DiaObjects", objects.shape[0])
700 return objects
702 def getDiaSources(
703 self, region: sphgeom.Region, object_ids: Iterable[int] | None, visit_time: astropy.time.Time
704 ) -> pandas.DataFrame | None:
705 # docstring is inherited from a base class
706 months = self.config.read_sources_months
707 if months == 0:
708 return None
709 mjd_end = visit_time.mjd
710 mjd_start = mjd_end - months * 30
712 return self._getSources(region, object_ids, mjd_start, mjd_end, ApdbTables.DiaSource)
714 def getDiaForcedSources(
715 self, region: sphgeom.Region, object_ids: Iterable[int] | None, visit_time: astropy.time.Time
716 ) -> pandas.DataFrame | None:
717 # docstring is inherited from a base class
718 months = self.config.read_forced_sources_months
719 if months == 0:
720 return None
721 mjd_end = visit_time.mjd
722 mjd_start = mjd_end - months * 30
724 return self._getSources(region, object_ids, mjd_start, mjd_end, ApdbTables.DiaForcedSource)
726 def containsVisitDetector(self, visit: int, detector: int) -> bool:
727 # docstring is inherited from a base class
728 raise NotImplementedError()
730 def getSSObjects(self) -> pandas.DataFrame:
731 # docstring is inherited from a base class
732 tableName = self._schema.tableName(ApdbTables.SSObject)
733 query = f'SELECT * from "{self._keyspace}"."{tableName}"'
735 objects = None
736 with self._timer("select_time", tags={"table": "SSObject"}):
737 result = self._session.execute(query, execution_profile="read_pandas")
738 objects = result._current_rows
740 _LOG.debug("found %s SSObjects", objects.shape[0])
741 return objects
743 def store(
744 self,
745 visit_time: astropy.time.Time,
746 objects: pandas.DataFrame,
747 sources: pandas.DataFrame | None = None,
748 forced_sources: pandas.DataFrame | None = None,
749 ) -> None:
750 # docstring is inherited from a base class
752 replica_chunk: ReplicaChunk | None = None
753 if self._schema.has_replica_chunks:
754 replica_chunk = ReplicaChunk.make_replica_chunk(visit_time, self.config.replica_chunk_seconds)
755 self._storeReplicaChunk(replica_chunk, visit_time)
757 # fill region partition column for DiaObjects
758 objects = self._add_obj_part(objects)
759 self._storeDiaObjects(objects, visit_time, replica_chunk)
761 if sources is not None:
762 # copy apdb_part column from DiaObjects to DiaSources
763 sources = self._add_src_part(sources, objects)
764 self._storeDiaSources(ApdbTables.DiaSource, sources, visit_time, replica_chunk)
765 self._storeDiaSourcesPartitions(sources, visit_time, replica_chunk)
767 if forced_sources is not None:
768 forced_sources = self._add_fsrc_part(forced_sources, objects)
769 self._storeDiaSources(ApdbTables.DiaForcedSource, forced_sources, visit_time, replica_chunk)
771 def storeSSObjects(self, objects: pandas.DataFrame) -> None:
772 # docstring is inherited from a base class
773 self._storeObjectsPandas(objects, ApdbTables.SSObject)
775 def reassignDiaSources(self, idMap: Mapping[int, int]) -> None:
776 # docstring is inherited from a base class
778 # To update a record we need to know its exact primary key (including
779 # partition key) so we start by querying for diaSourceId to find the
780 # primary keys.
782 table_name = self._schema.tableName(ExtraTables.DiaSourceToPartition)
783 # split it into 1k IDs per query
784 selects: list[tuple] = []
785 for ids in chunk_iterable(idMap.keys(), 1_000):
786 ids_str = ",".join(str(item) for item in ids)
787 selects.append(
788 (
789 (
790 'SELECT "diaSourceId", "apdb_part", "apdb_time_part", "apdb_replica_chunk" '
791 f'FROM "{self._keyspace}"."{table_name}" WHERE "diaSourceId" IN ({ids_str})'
792 ),
793 {},
794 )
795 )
797 # No need for DataFrame here, read data as tuples.
798 result = cast(
799 list[tuple[int, int, int, int | None]],
800 select_concurrent(self._session, selects, "read_tuples", self.config.read_concurrency),
801 )
803 # Make mapping from source ID to its partition.
804 id2partitions: dict[int, tuple[int, int]] = {}
805 id2chunk_id: dict[int, int] = {}
806 for row in result:
807 id2partitions[row[0]] = row[1:3]
808 if row[3] is not None:
809 id2chunk_id[row[0]] = row[3]
811 # make sure we know partitions for each ID
812 if set(id2partitions) != set(idMap):
813 missing = ",".join(str(item) for item in set(idMap) - set(id2partitions))
814 raise ValueError(f"Following DiaSource IDs do not exist in the database: {missing}")
816 # Reassign in standard tables
817 queries = cassandra.query.BatchStatement()
818 table_name = self._schema.tableName(ApdbTables.DiaSource)
819 for diaSourceId, ssObjectId in idMap.items():
820 apdb_part, apdb_time_part = id2partitions[diaSourceId]
821 values: tuple
822 if self.config.time_partition_tables:
823 query = (
824 f'UPDATE "{self._keyspace}"."{table_name}_{apdb_time_part}"'
825 ' SET "ssObjectId" = ?, "diaObjectId" = NULL'
826 ' WHERE "apdb_part" = ? AND "diaSourceId" = ?'
827 )
828 values = (ssObjectId, apdb_part, diaSourceId)
829 else:
830 query = (
831 f'UPDATE "{self._keyspace}"."{table_name}"'
832 ' SET "ssObjectId" = ?, "diaObjectId" = NULL'
833 ' WHERE "apdb_part" = ? AND "apdb_time_part" = ? AND "diaSourceId" = ?'
834 )
835 values = (ssObjectId, apdb_part, apdb_time_part, diaSourceId)
836 queries.add(self._preparer.prepare(query), values)
838 # Reassign in replica tables, only if replication is enabled
839 if id2chunk_id:
840 # Filter out chunks that have been deleted already. There is a
841 # potential race with concurrent removal of chunks, but it
842 # should be handled by WHERE in UPDATE.
843 known_ids = set()
844 if replica_chunks := self.get_replica().getReplicaChunks():
845 known_ids = set(replica_chunk.id for replica_chunk in replica_chunks)
846 id2chunk_id = {key: value for key, value in id2chunk_id.items() if value in known_ids}
847 if id2chunk_id:
848 table_name = self._schema.tableName(ExtraTables.DiaSourceChunks)
849 for diaSourceId, ssObjectId in idMap.items():
850 if replica_chunk := id2chunk_id.get(diaSourceId):
851 query = (
852 f'UPDATE "{self._keyspace}"."{table_name}" '
853 ' SET "ssObjectId" = ?, "diaObjectId" = NULL '
854 'WHERE "apdb_replica_chunk" = ? AND "diaSourceId" = ?'
855 )
856 values = (ssObjectId, replica_chunk, diaSourceId)
857 queries.add(self._preparer.prepare(query), values)
859 _LOG.debug("%s: will update %d records", table_name, len(idMap))
860 with self._timer("update_time", tags={"table": table_name}):
861 self._session.execute(queries, execution_profile="write")
863 def dailyJob(self) -> None:
864 # docstring is inherited from a base class
865 pass
867 def countUnassociatedObjects(self) -> int:
868 # docstring is inherited from a base class
870 # It's too inefficient to implement it for Cassandra in current schema.
871 raise NotImplementedError()
873 @property
874 def metadata(self) -> ApdbMetadata:
875 # docstring is inherited from a base class
876 if self._metadata is None:
877 raise RuntimeError("Database schema was not initialized.")
878 return self._metadata
880 @classmethod
881 def _makeProfiles(cls, config: ApdbCassandraConfig) -> Mapping[Any, ExecutionProfile]:
882 """Make all execution profiles used in the code."""
883 if config.private_ips:
884 loadBalancePolicy = WhiteListRoundRobinPolicy(hosts=config.contact_points)
885 else:
886 loadBalancePolicy = RoundRobinPolicy()
888 read_tuples_profile = ExecutionProfile(
889 consistency_level=getattr(cassandra.ConsistencyLevel, config.read_consistency),
890 request_timeout=config.read_timeout,
891 row_factory=cassandra.query.tuple_factory,
892 load_balancing_policy=loadBalancePolicy,
893 )
894 read_pandas_profile = ExecutionProfile(
895 consistency_level=getattr(cassandra.ConsistencyLevel, config.read_consistency),
896 request_timeout=config.read_timeout,
897 row_factory=pandas_dataframe_factory,
898 load_balancing_policy=loadBalancePolicy,
899 )
900 read_raw_profile = ExecutionProfile(
901 consistency_level=getattr(cassandra.ConsistencyLevel, config.read_consistency),
902 request_timeout=config.read_timeout,
903 row_factory=raw_data_factory,
904 load_balancing_policy=loadBalancePolicy,
905 )
906 # Profile to use with select_concurrent to return pandas data frame
907 read_pandas_multi_profile = ExecutionProfile(
908 consistency_level=getattr(cassandra.ConsistencyLevel, config.read_consistency),
909 request_timeout=config.read_timeout,
910 row_factory=pandas_dataframe_factory,
911 load_balancing_policy=loadBalancePolicy,
912 )
913 # Profile to use with select_concurrent to return raw data (columns and
914 # rows)
915 read_raw_multi_profile = ExecutionProfile(
916 consistency_level=getattr(cassandra.ConsistencyLevel, config.read_consistency),
917 request_timeout=config.read_timeout,
918 row_factory=raw_data_factory,
919 load_balancing_policy=loadBalancePolicy,
920 )
921 write_profile = ExecutionProfile(
922 consistency_level=getattr(cassandra.ConsistencyLevel, config.write_consistency),
923 request_timeout=config.write_timeout,
924 load_balancing_policy=loadBalancePolicy,
925 )
926 # To replace default DCAwareRoundRobinPolicy
927 default_profile = ExecutionProfile(
928 load_balancing_policy=loadBalancePolicy,
929 )
930 return {
931 "read_tuples": read_tuples_profile,
932 "read_pandas": read_pandas_profile,
933 "read_raw": read_raw_profile,
934 "read_pandas_multi": read_pandas_multi_profile,
935 "read_raw_multi": read_raw_multi_profile,
936 "write": write_profile,
937 EXEC_PROFILE_DEFAULT: default_profile,
938 }
940 def _getSources(
941 self,
942 region: sphgeom.Region,
943 object_ids: Iterable[int] | None,
944 mjd_start: float,
945 mjd_end: float,
946 table_name: ApdbTables,
947 ) -> pandas.DataFrame:
948 """Return catalog of DiaSource instances given set of DiaObject IDs.
950 Parameters
951 ----------
952 region : `lsst.sphgeom.Region`
953 Spherical region.
954 object_ids :
955 Collection of DiaObject IDs
956 mjd_start : `float`
957 Lower bound of time interval.
958 mjd_end : `float`
959 Upper bound of time interval.
960 table_name : `ApdbTables`
961 Name of the table.
963 Returns
964 -------
965 catalog : `pandas.DataFrame`, or `None`
966 Catalog containing DiaSource records. Empty catalog is returned if
967 ``object_ids`` is empty.
968 """
969 object_id_set: Set[int] = set()
970 if object_ids is not None:
971 object_id_set = set(object_ids)
972 if len(object_id_set) == 0:
973 return self._make_empty_catalog(table_name)
975 sp_where = self._spatial_where(region)
976 tables, temporal_where = self._temporal_where(table_name, mjd_start, mjd_end)
978 # We need to exclude extra partitioning columns from result.
979 column_names = self._schema.apdbColumnNames(table_name)
980 what = ",".join(quote_id(column) for column in column_names)
982 # Build all queries
983 statements: list[tuple] = []
984 for table in tables:
985 prefix = f'SELECT {what} from "{self._keyspace}"."{table}"'
986 statements += list(self._combine_where(prefix, sp_where, temporal_where))
987 _LOG.debug("_getSources %s: #queries: %s", table_name, len(statements))
989 with _MON.context_tags({"table": table_name.name}):
990 _MON.add_record(
991 "select_query_stats", values={"num_sp_part": len(sp_where), "num_queries": len(statements)}
992 )
993 with self._timer("select_time"):
994 catalog = cast(
995 pandas.DataFrame,
996 select_concurrent(
997 self._session, statements, "read_pandas_multi", self.config.read_concurrency
998 ),
999 )
1001 # filter by given object IDs
1002 if len(object_id_set) > 0:
1003 catalog = cast(pandas.DataFrame, catalog[catalog["diaObjectId"].isin(object_id_set)])
1005 # precise filtering on midpointMjdTai
1006 catalog = cast(pandas.DataFrame, catalog[catalog["midpointMjdTai"] > mjd_start])
1008 _LOG.debug("found %d %ss", catalog.shape[0], table_name.name)
1009 return catalog
1011 def _storeReplicaChunk(self, replica_chunk: ReplicaChunk, visit_time: astropy.time.Time) -> None:
1012 # Cassandra timestamp uses milliseconds since epoch
1013 timestamp = int(replica_chunk.last_update_time.unix_tai * 1000)
1015 # everything goes into a single partition
1016 partition = 0
1018 table_name = self._schema.tableName(ExtraTables.ApdbReplicaChunks)
1019 query = (
1020 f'INSERT INTO "{self._keyspace}"."{table_name}" '
1021 "(partition, apdb_replica_chunk, last_update_time, unique_id) "
1022 "VALUES (?, ?, ?, ?)"
1023 )
1025 self._session.execute(
1026 self._preparer.prepare(query),
1027 (partition, replica_chunk.id, timestamp, replica_chunk.unique_id),
1028 timeout=self.config.write_timeout,
1029 execution_profile="write",
1030 )
1032 def _storeDiaObjects(
1033 self, objs: pandas.DataFrame, visit_time: astropy.time.Time, replica_chunk: ReplicaChunk | None
1034 ) -> None:
1035 """Store catalog of DiaObjects from current visit.
1037 Parameters
1038 ----------
1039 objs : `pandas.DataFrame`
1040 Catalog with DiaObject records
1041 visit_time : `astropy.time.Time`
1042 Time of the current visit.
1043 replica_chunk : `ReplicaChunk` or `None`
1044 Replica chunk identifier if replication is configured.
1045 """
1046 if len(objs) == 0:
1047 _LOG.debug("No objects to write to database.")
1048 return
1050 visit_time_dt = visit_time.datetime
1051 extra_columns = dict(lastNonForcedSource=visit_time_dt)
1052 self._storeObjectsPandas(objs, ApdbTables.DiaObjectLast, extra_columns=extra_columns)
1054 extra_columns["validityStart"] = visit_time_dt
1055 time_part: int | None = self._time_partition(visit_time)
1056 if not self.config.time_partition_tables:
1057 extra_columns["apdb_time_part"] = time_part
1058 time_part = None
1060 # Only store DiaObects if not doing replication or explicitly
1061 # configured to always store them.
1062 if replica_chunk is None or not self.config.use_insert_id_skips_diaobjects:
1063 self._storeObjectsPandas(
1064 objs, ApdbTables.DiaObject, extra_columns=extra_columns, time_part=time_part
1065 )
1067 if replica_chunk is not None:
1068 extra_columns = dict(apdb_replica_chunk=replica_chunk.id, validityStart=visit_time_dt)
1069 self._storeObjectsPandas(objs, ExtraTables.DiaObjectChunks, extra_columns=extra_columns)
1071 def _storeDiaSources(
1072 self,
1073 table_name: ApdbTables,
1074 sources: pandas.DataFrame,
1075 visit_time: astropy.time.Time,
1076 replica_chunk: ReplicaChunk | None,
1077 ) -> None:
1078 """Store catalog of DIASources or DIAForcedSources from current visit.
1080 Parameters
1081 ----------
1082 table_name : `ApdbTables`
1083 Table where to store the data.
1084 sources : `pandas.DataFrame`
1085 Catalog containing DiaSource records
1086 visit_time : `astropy.time.Time`
1087 Time of the current visit.
1088 replica_chunk : `ReplicaChunk` or `None`
1089 Replica chunk identifier if replication is configured.
1090 """
1091 time_part: int | None = self._time_partition(visit_time)
1092 extra_columns: dict[str, Any] = {}
1093 if not self.config.time_partition_tables:
1094 extra_columns["apdb_time_part"] = time_part
1095 time_part = None
1097 self._storeObjectsPandas(sources, table_name, extra_columns=extra_columns, time_part=time_part)
1099 if replica_chunk is not None:
1100 extra_columns = dict(apdb_replica_chunk=replica_chunk.id)
1101 if table_name is ApdbTables.DiaSource:
1102 extra_table = ExtraTables.DiaSourceChunks
1103 else:
1104 extra_table = ExtraTables.DiaForcedSourceChunks
1105 self._storeObjectsPandas(sources, extra_table, extra_columns=extra_columns)
1107 def _storeDiaSourcesPartitions(
1108 self, sources: pandas.DataFrame, visit_time: astropy.time.Time, replica_chunk: ReplicaChunk | None
1109 ) -> None:
1110 """Store mapping of diaSourceId to its partitioning values.
1112 Parameters
1113 ----------
1114 sources : `pandas.DataFrame`
1115 Catalog containing DiaSource records
1116 visit_time : `astropy.time.Time`
1117 Time of the current visit.
1118 """
1119 id_map = cast(pandas.DataFrame, sources[["diaSourceId", "apdb_part"]])
1120 extra_columns = {
1121 "apdb_time_part": self._time_partition(visit_time),
1122 "apdb_replica_chunk": replica_chunk.id if replica_chunk is not None else None,
1123 }
1125 self._storeObjectsPandas(
1126 id_map, ExtraTables.DiaSourceToPartition, extra_columns=extra_columns, time_part=None
1127 )
1129 def _storeObjectsPandas(
1130 self,
1131 records: pandas.DataFrame,
1132 table_name: ApdbTables | ExtraTables,
1133 extra_columns: Mapping | None = None,
1134 time_part: int | None = None,
1135 ) -> None:
1136 """Store generic objects.
1138 Takes Pandas catalog and stores a bunch of records in a table.
1140 Parameters
1141 ----------
1142 records : `pandas.DataFrame`
1143 Catalog containing object records
1144 table_name : `ApdbTables`
1145 Name of the table as defined in APDB schema.
1146 extra_columns : `dict`, optional
1147 Mapping (column_name, column_value) which gives fixed values for
1148 columns in each row, overrides values in ``records`` if matching
1149 columns exist there.
1150 time_part : `int`, optional
1151 If not `None` then insert into a per-partition table.
1153 Notes
1154 -----
1155 If Pandas catalog contains additional columns not defined in table
1156 schema they are ignored. Catalog does not have to contain all columns
1157 defined in a table, but partition and clustering keys must be present
1158 in a catalog or ``extra_columns``.
1159 """
1160 # use extra columns if specified
1161 if extra_columns is None:
1162 extra_columns = {}
1163 extra_fields = list(extra_columns.keys())
1165 # Fields that will come from dataframe.
1166 df_fields = [column for column in records.columns if column not in extra_fields]
1168 column_map = self._schema.getColumnMap(table_name)
1169 # list of columns (as in felis schema)
1170 fields = [column_map[field].name for field in df_fields if field in column_map]
1171 fields += extra_fields
1173 # check that all partitioning and clustering columns are defined
1174 required_columns = self._schema.partitionColumns(table_name) + self._schema.clusteringColumns(
1175 table_name
1176 )
1177 missing_columns = [column for column in required_columns if column not in fields]
1178 if missing_columns:
1179 raise ValueError(f"Primary key columns are missing from catalog: {missing_columns}")
1181 qfields = [quote_id(field) for field in fields]
1182 qfields_str = ",".join(qfields)
1184 with self._timer("insert_build_time", tags={"table": table_name.name}):
1185 table = self._schema.tableName(table_name)
1186 if time_part is not None:
1187 table = f"{table}_{time_part}"
1189 holders = ",".join(["?"] * len(qfields))
1190 query = f'INSERT INTO "{self._keyspace}"."{table}" ({qfields_str}) VALUES ({holders})'
1191 statement = self._preparer.prepare(query)
1192 queries = cassandra.query.BatchStatement()
1193 for rec in records.itertuples(index=False):
1194 values = []
1195 for field in df_fields:
1196 if field not in column_map:
1197 continue
1198 value = getattr(rec, field)
1199 if column_map[field].datatype is felis.datamodel.DataType.timestamp:
1200 if isinstance(value, pandas.Timestamp):
1201 value = literal(value.to_pydatetime())
1202 else:
1203 # Assume it's seconds since epoch, Cassandra
1204 # datetime is in milliseconds
1205 value = int(value * 1000)
1206 values.append(literal(value))
1207 for field in extra_fields:
1208 value = extra_columns[field]
1209 values.append(literal(value))
1210 queries.add(statement, values)
1212 _LOG.debug("%s: will store %d records", self._schema.tableName(table_name), records.shape[0])
1213 with self._timer("insert_time", tags={"table": table_name.name}):
1214 self._session.execute(queries, timeout=self.config.write_timeout, execution_profile="write")
1216 def _add_obj_part(self, df: pandas.DataFrame) -> pandas.DataFrame:
1217 """Calculate spatial partition for each record and add it to a
1218 DataFrame.
1220 Notes
1221 -----
1222 This overrides any existing column in a DataFrame with the same name
1223 (apdb_part). Original DataFrame is not changed, copy of a DataFrame is
1224 returned.
1225 """
1226 # calculate HTM index for every DiaObject
1227 apdb_part = np.zeros(df.shape[0], dtype=np.int64)
1228 ra_col, dec_col = self.config.ra_dec_columns
1229 for i, (ra, dec) in enumerate(zip(df[ra_col], df[dec_col])):
1230 uv3d = sphgeom.UnitVector3d(sphgeom.LonLat.fromDegrees(ra, dec))
1231 idx = self._pixelization.pixel(uv3d)
1232 apdb_part[i] = idx
1233 df = df.copy()
1234 df["apdb_part"] = apdb_part
1235 return df
1237 def _add_src_part(self, sources: pandas.DataFrame, objs: pandas.DataFrame) -> pandas.DataFrame:
1238 """Add apdb_part column to DiaSource catalog.
1240 Notes
1241 -----
1242 This method copies apdb_part value from a matching DiaObject record.
1243 DiaObject catalog needs to have a apdb_part column filled by
1244 ``_add_obj_part`` method and DiaSource records need to be
1245 associated to DiaObjects via ``diaObjectId`` column.
1247 This overrides any existing column in a DataFrame with the same name
1248 (apdb_part). Original DataFrame is not changed, copy of a DataFrame is
1249 returned.
1250 """
1251 pixel_id_map: dict[int, int] = {
1252 diaObjectId: apdb_part for diaObjectId, apdb_part in zip(objs["diaObjectId"], objs["apdb_part"])
1253 }
1254 apdb_part = np.zeros(sources.shape[0], dtype=np.int64)
1255 ra_col, dec_col = self.config.ra_dec_columns
1256 for i, (diaObjId, ra, dec) in enumerate(
1257 zip(sources["diaObjectId"], sources[ra_col], sources[dec_col])
1258 ):
1259 if diaObjId == 0:
1260 # DiaSources associated with SolarSystemObjects do not have an
1261 # associated DiaObject hence we skip them and set partition
1262 # based on its own ra/dec
1263 uv3d = sphgeom.UnitVector3d(sphgeom.LonLat.fromDegrees(ra, dec))
1264 idx = self._pixelization.pixel(uv3d)
1265 apdb_part[i] = idx
1266 else:
1267 apdb_part[i] = pixel_id_map[diaObjId]
1268 sources = sources.copy()
1269 sources["apdb_part"] = apdb_part
1270 return sources
1272 def _add_fsrc_part(self, sources: pandas.DataFrame, objs: pandas.DataFrame) -> pandas.DataFrame:
1273 """Add apdb_part column to DiaForcedSource catalog.
1275 Notes
1276 -----
1277 This method copies apdb_part value from a matching DiaObject record.
1278 DiaObject catalog needs to have a apdb_part column filled by
1279 ``_add_obj_part`` method and DiaSource records need to be
1280 associated to DiaObjects via ``diaObjectId`` column.
1282 This overrides any existing column in a DataFrame with the same name
1283 (apdb_part). Original DataFrame is not changed, copy of a DataFrame is
1284 returned.
1285 """
1286 pixel_id_map: dict[int, int] = {
1287 diaObjectId: apdb_part for diaObjectId, apdb_part in zip(objs["diaObjectId"], objs["apdb_part"])
1288 }
1289 apdb_part = np.zeros(sources.shape[0], dtype=np.int64)
1290 for i, diaObjId in enumerate(sources["diaObjectId"]):
1291 apdb_part[i] = pixel_id_map[diaObjId]
1292 sources = sources.copy()
1293 sources["apdb_part"] = apdb_part
1294 return sources
1296 @classmethod
1297 def _time_partition_cls(cls, time: float | astropy.time.Time, epoch_mjd: float, part_days: int) -> int:
1298 """Calculate time partition number for a given time.
1300 Parameters
1301 ----------
1302 time : `float` or `astropy.time.Time`
1303 Time for which to calculate partition number. Can be float to mean
1304 MJD or `astropy.time.Time`
1305 epoch_mjd : `float`
1306 Epoch time for partition 0.
1307 part_days : `int`
1308 Number of days per partition.
1310 Returns
1311 -------
1312 partition : `int`
1313 Partition number for a given time.
1314 """
1315 if isinstance(time, astropy.time.Time):
1316 mjd = float(time.mjd)
1317 else:
1318 mjd = time
1319 days_since_epoch = mjd - epoch_mjd
1320 partition = int(days_since_epoch) // part_days
1321 return partition
1323 def _time_partition(self, time: float | astropy.time.Time) -> int:
1324 """Calculate time partition number for a given time.
1326 Parameters
1327 ----------
1328 time : `float` or `astropy.time.Time`
1329 Time for which to calculate partition number. Can be float to mean
1330 MJD or `astropy.time.Time`
1332 Returns
1333 -------
1334 partition : `int`
1335 Partition number for a given time.
1336 """
1337 if isinstance(time, astropy.time.Time):
1338 mjd = float(time.mjd)
1339 else:
1340 mjd = time
1341 days_since_epoch = mjd - self._partition_zero_epoch_mjd
1342 partition = int(days_since_epoch) // self.config.time_partition_days
1343 return partition
1345 def _make_empty_catalog(self, table_name: ApdbTables) -> pandas.DataFrame:
1346 """Make an empty catalog for a table with a given name.
1348 Parameters
1349 ----------
1350 table_name : `ApdbTables`
1351 Name of the table.
1353 Returns
1354 -------
1355 catalog : `pandas.DataFrame`
1356 An empty catalog.
1357 """
1358 table = self._schema.tableSchemas[table_name]
1360 data = {
1361 columnDef.name: pandas.Series(dtype=self._schema.column_dtype(columnDef.datatype))
1362 for columnDef in table.columns
1363 }
1364 return pandas.DataFrame(data)
1366 def _combine_where(
1367 self,
1368 prefix: str,
1369 where1: list[tuple[str, tuple]],
1370 where2: list[tuple[str, tuple]],
1371 suffix: str | None = None,
1372 ) -> Iterator[tuple[cassandra.query.Statement, tuple]]:
1373 """Make cartesian product of two parts of WHERE clause into a series
1374 of statements to execute.
1376 Parameters
1377 ----------
1378 prefix : `str`
1379 Initial statement prefix that comes before WHERE clause, e.g.
1380 "SELECT * from Table"
1381 """
1382 # If lists are empty use special sentinels.
1383 if not where1:
1384 where1 = [("", ())]
1385 if not where2:
1386 where2 = [("", ())]
1388 for expr1, params1 in where1:
1389 for expr2, params2 in where2:
1390 full_query = prefix
1391 wheres = []
1392 if expr1:
1393 wheres.append(expr1)
1394 if expr2:
1395 wheres.append(expr2)
1396 if wheres:
1397 full_query += " WHERE " + " AND ".join(wheres)
1398 if suffix:
1399 full_query += " " + suffix
1400 params = params1 + params2
1401 if params:
1402 statement = self._preparer.prepare(full_query)
1403 else:
1404 # If there are no params then it is likely that query
1405 # has a bunch of literals rendered already, no point
1406 # trying to prepare it.
1407 statement = cassandra.query.SimpleStatement(full_query)
1408 yield (statement, params)
1410 def _spatial_where(
1411 self, region: sphgeom.Region | None, use_ranges: bool = False
1412 ) -> list[tuple[str, tuple]]:
1413 """Generate expressions for spatial part of WHERE clause.
1415 Parameters
1416 ----------
1417 region : `sphgeom.Region`
1418 Spatial region for query results.
1419 use_ranges : `bool`
1420 If True then use pixel ranges ("apdb_part >= p1 AND apdb_part <=
1421 p2") instead of exact list of pixels. Should be set to True for
1422 large regions covering very many pixels.
1424 Returns
1425 -------
1426 expressions : `list` [ `tuple` ]
1427 Empty list is returned if ``region`` is `None`, otherwise a list
1428 of one or more (expression, parameters) tuples
1429 """
1430 if region is None:
1431 return []
1432 if use_ranges:
1433 pixel_ranges = self._pixelization.envelope(region)
1434 expressions: list[tuple[str, tuple]] = []
1435 for lower, upper in pixel_ranges:
1436 upper -= 1
1437 if lower == upper:
1438 expressions.append(('"apdb_part" = ?', (lower,)))
1439 else:
1440 expressions.append(('"apdb_part" >= ? AND "apdb_part" <= ?', (lower, upper)))
1441 return expressions
1442 else:
1443 pixels = self._pixelization.pixels(region)
1444 if self.config.query_per_spatial_part:
1445 return [('"apdb_part" = ?', (pixel,)) for pixel in pixels]
1446 else:
1447 pixels_str = ",".join([str(pix) for pix in pixels])
1448 return [(f'"apdb_part" IN ({pixels_str})', ())]
1450 def _temporal_where(
1451 self,
1452 table: ApdbTables,
1453 start_time: float | astropy.time.Time,
1454 end_time: float | astropy.time.Time,
1455 query_per_time_part: bool | None = None,
1456 ) -> tuple[list[str], list[tuple[str, tuple]]]:
1457 """Generate table names and expressions for temporal part of WHERE
1458 clauses.
1460 Parameters
1461 ----------
1462 table : `ApdbTables`
1463 Table to select from.
1464 start_time : `astropy.time.Time` or `float`
1465 Starting Datetime of MJD value of the time range.
1466 end_time : `astropy.time.Time` or `float`
1467 Starting Datetime of MJD value of the time range.
1468 query_per_time_part : `bool`, optional
1469 If None then use ``query_per_time_part`` from configuration.
1471 Returns
1472 -------
1473 tables : `list` [ `str` ]
1474 List of the table names to query.
1475 expressions : `list` [ `tuple` ]
1476 A list of zero or more (expression, parameters) tuples.
1477 """
1478 tables: list[str]
1479 temporal_where: list[tuple[str, tuple]] = []
1480 table_name = self._schema.tableName(table)
1481 time_part_start = self._time_partition(start_time)
1482 time_part_end = self._time_partition(end_time)
1483 time_parts = list(range(time_part_start, time_part_end + 1))
1484 if self.config.time_partition_tables:
1485 tables = [f"{table_name}_{part}" for part in time_parts]
1486 else:
1487 tables = [table_name]
1488 if query_per_time_part is None:
1489 query_per_time_part = self.config.query_per_time_part
1490 if query_per_time_part:
1491 temporal_where = [('"apdb_time_part" = ?', (time_part,)) for time_part in time_parts]
1492 else:
1493 time_part_list = ",".join([str(part) for part in time_parts])
1494 temporal_where = [(f'"apdb_time_part" IN ({time_part_list})', ())]
1496 return tables, temporal_where