Remove hydra use query api (#491)

cyclops/query/base.py

@@ -1,25 +1,20 @@
 """Base querier class."""
 
 import logging
-import os
 from functools import partial
 from typing import Any, Callable, Dict, List, Optional
 
-import yaml
-from hydra import compose, initialize
-from omegaconf import OmegaConf
 from sqlalchemy import MetaData
 from sqlalchemy.sql.selectable import Subquery
 
 from cyclops.query import ops as qo
 from cyclops.query.interface import QueryInterface
-from cyclops.query.orm import Database
+from cyclops.query.orm import Database, DatasetQuerierConfig
 from cyclops.query.util import (
     DBSchema,
     _to_subquery,
     get_attr_name,
 )
-from cyclops.utils.file import join as join_path
 from cyclops.utils.log import setup_logging
 
 
@@ -81,9 +76,18 @@ class DatasetQuerier:
 
     Parameters
     ----------
-    config_overrides
-        Override configuration parameters, specified as kwargs.
-
+    database
+        Name of database.
+    user
+        Username for database.
+    password
+        Password for database.
+    dbms
+        Database management system.
+    host
+        Hostname of database.
+    port
+        Port of database.
 
     Notes
     -----
@@ -98,26 +102,21 @@ class DatasetQuerier:
 
     def __init__(
         self,
-        **config_overrides: Dict[str, Any],
+        database: str,
+        user: str,
+        password: str,
+        dbms: str = "postgresql",
+        host: str = "localhost",
+        port: int = 5432,
     ) -> None:
-        overrides = []
-        if config_overrides:
-            config_file = join_path(os.path.dirname(__file__), "configs", "config.yaml")
-            with open(config_file, "r", encoding="utf-8") as file:
-                config_keys = list(yaml.safe_load(file).keys())
-            for key, value in config_overrides.items():
-                if key in config_keys:
-                    overrides.append(f"{key}={value}")
-                else:
-                    overrides.append(f"+{key}={value}")
-        with initialize(
-            version_base=None,
-            config_path="configs",
-            job_name="DatasetQuerier",
-        ):
-            config = compose(config_name="config", overrides=overrides)
-            LOGGER.debug(OmegaConf.to_yaml(config))
-
+        config = DatasetQuerierConfig(
+            database=database,
+            user=user,
+            password=password,
+            dbms=dbms,
+            host=host,
+            port=port,
+        )
         self.db = Database(config)
         if not self.db.is_connected:
             LOGGER.error("Database is not connected, cannot run queries.")

cyclops/query/eicu.py

@@ -5,7 +5,6 @@
 """
 
 import logging
-from typing import Any, Dict
 
 from cyclops.query.base import DatasetQuerier
 from cyclops.utils.log import setup_logging
@@ -18,17 +17,3 @@
 
 class EICUQuerier(DatasetQuerier):
     """EICU dataset querier."""
-
-    def __init__(self, **config_overrides: Dict[str, Any]) -> None:
-        """Initialize.
-
-        Parameters
-        ----------
-        **config_overrides
-            Override configuration parameters, specified as kwargs.
-
-        """
-        overrides = {}
-        if config_overrides:
-            overrides = config_overrides
-        super().__init__(**overrides)

cyclops/query/gemini.py

@@ -1,7 +1,6 @@
 """GEMINI query API."""
 
 import logging
-from typing import Any, Dict
 
 from sqlalchemy import select
 from sqlalchemy.sql.expression import union_all
@@ -24,20 +23,6 @@
 class GEMINIQuerier(DatasetQuerier):
     """GEMINI dataset querier."""
 
-    def __init__(self, **config_overrides: Dict[str, Any]) -> None:
-        """Initialize.
-
-        Parameters
-        ----------
-        **config_overrides
-            Override configuration parameters, specified as kwargs.
-
-        """
-        overrides = {}
-        if config_overrides:
-            overrides = config_overrides
-        super().__init__(**overrides)
-
     def ip_admin(
         self,
     ) -> QueryInterface:

cyclops/query/mimiciii.py

@@ -5,7 +5,6 @@
 """
 
 import logging
-from typing import Any, Dict
 
 import cyclops.query.ops as qo
 from cyclops.query.base import DatasetQuerier
@@ -21,20 +20,6 @@
 class MIMICIIIQuerier(DatasetQuerier):
     """MIMIC-III dataset querier."""
 
-    def __init__(self, **config_overrides: Dict[str, Any]) -> None:
-        """Initialize.
-
-        Parameters
-        ----------
-        **config_overrides
-            Override configuration parameters, specified as kwargs.
-
-        """
-        overrides = {}
-        if config_overrides:
-            overrides = config_overrides
-        super().__init__(**overrides)
-
     def diagnoses(
         self,
     ) -> QueryInterface:

cyclops/query/mimiciv.py

@@ -5,7 +5,6 @@
 """
 
 import logging
-from typing import Any, Dict
 
 from sqlalchemy import Integer, func, select
 
@@ -24,20 +23,6 @@
 class MIMICIVQuerier(DatasetQuerier):
     """MIMICIV dataset querier."""
 
-    def __init__(self, **config_overrides: Dict[str, Any]) -> None:
-        """Initialize.
-
-        Parameters
-        ----------
-        **config_overrides
-            Override configuration parameters, specified as kwargs.
-
-        """
-        overrides = {}
-        if config_overrides:
-            overrides = config_overrides
-        super().__init__(**overrides)
-
     def patients(
         self,
     ) -> QueryInterface:

cyclops/query/omop.py

@@ -1,7 +1,7 @@
 """OMOP query API."""
 
 import logging
-from typing import Any, Dict, List, Optional, Union
+from typing import List, Optional, Union
 
 from sqlalchemy.sql.selectable import Subquery
 
@@ -58,24 +58,23 @@ class OMOPQuerier(DatasetQuerier):
 
     def __init__(
         self,
-        schema_name: str,
-        **config_overrides: Dict[str, Any],
+        database: str,
+        user: str,
+        password: str,
+        dbms: str = "postgresql",
+        host: str = "localhost",
+        port: int = 5432,
+        schema_name: str = "omop",
     ) -> None:
-        """Initialize.
-
-        Parameters
-        ----------
-        schema_name
-            Name of database schema.
-        **config_overrides
-            Override configuration parameters, specified as kwargs.
-
-        """
+        super().__init__(
+            database=database,
+            user=user,
+            password=password,
+            dbms=dbms,
+            host=host,
+            port=port,
+        )
         self.schema_name = schema_name
-        overrides = {}
-        if config_overrides:
-            overrides = config_overrides
-        super().__init__(**overrides)
 
     def map_concept_ids_to_name(
         self,

cyclops/query/orm.py

@@ -4,14 +4,15 @@
 import logging
 import os
 import socket
+from dataclasses import dataclass
 from typing import Dict, List, Literal, Optional, Union
+from urllib.parse import quote_plus
 
 import dask.dataframe as dd
 import pandas as pd
 import pyarrow.csv as pv
 import pyarrow.parquet as pq
 from datasets import Dataset
-from omegaconf import DictConfig
 from sqlalchemy import MetaData, create_engine, inspect
 from sqlalchemy.engine.base import Engine
 from sqlalchemy.orm import sessionmaker
@@ -43,32 +44,61 @@ def _get_db_url(
     user: str,
     pwd: str,
     host: str,
-    port: str,
+    port: int,
     database: str,
 ) -> str:
     """Combine to make Database URL string."""
-    return f"{dbms}://{user}:{pwd}@{host}:{port}/{database}"
+    return f"{dbms}://{user}:{quote_plus(pwd)}@{host}:{str(port)}/{database}"
+
+
+@dataclass
+class DatasetQuerierConfig:
+    """Configuration for the dataset querier.
+
+    Attributes
+    ----------
+    dbms
+        Database management system.
+    host
+        Hostname of database.
+    port
+        Port of database.
+    database
+        Name of database.
+    user
+        Username for database.
+    password
+        Password for database.
+
+    """
+
+    database: str
+    user: str
+    password: str
+    dbms: str = "postgresql"
+    host: str = "localhost"
+    port: int = 5432
 
 
 class Database:
     """Database class.
 
     Attributes
     ----------
-    config: argparse.Namespace
-        Configuration stored in object.
-    engine: sqlalchemy.engine.base.Engine
+    config
+        Configuration stored in a dataclass.
+    engine
         SQL extraction engine.
-    inspector: sqlalchemy.engine.reflection.Inspector
+    inspector
         Module for schema inspection.
-    session: sqlalchemy.orm.session.Session
+    session
         Session for ORM.
-    is_connected: bool
+    is_connected
         Whether the database is setup, connected and ready to run queries.
 
     """
 
-    def __init__(self, config: DictConfig) -> None:
+    def __init__(self, config: DatasetQuerierConfig) -> None:
         """Instantiate.
 
         Parameters