New Manybody Implementation

qcarchivetesting/conda-envs/fulltest_server.yaml

@@ -38,6 +38,7 @@ dependencies:
 
   # QCFractal Services
   - torsiondrive
+  - qcmanybody
 
   - pip:
       - "geometric @ git+https://github.com/leeping/geomeTRIC"

qcarchivetesting/conda-envs/fulltest_snowflake.yaml

@@ -41,6 +41,7 @@ dependencies:
 
   # QCFractal Services
   - torsiondrive
+  - qcmanybody
 
   # Worker codes below
   - qcengine<0.70a0

qcarchivetesting/qcarchivetesting/procedure_data/generate_manybody.py

@@ -35,7 +35,8 @@
 _, ids = client.add_manybodys(
     [molecule],
     program=test_data["specification"]["program"],
-    singlepoint_specification=test_data["specification"]["singlepoint_specification"],
+    bsse_correction=test_data["specification"]["bsse_correction"],
+    levels=test_data["specification"]["levels"],
     keywords=test_data["specification"]["keywords"],
 )
 

qcarchivetesting/qcarchivetesting/test_full_manybody.py

@@ -25,10 +25,14 @@ def test_manybody_full_1(fulltest_client: PortalClient):
         "keywords": {"e_convergence": 1e-10, "d_convergence": 1e-10},
     }
 
-    mb_keywords = {"max_nbody": None, "bsse_correction": "none"}
+    levels = {1: sp_spec, 2: sp_spec, 3: sp_spec, 4: sp_spec}
 
     meta, ids = fulltest_client.add_manybodys(
-        initial_molecules=[molecule], program="manybody", singlepoint_specification=sp_spec, keywords=mb_keywords
+        initial_molecules=[molecule],
+        program="qcmanybody",
+        bsse_correction=["nocp"],
+        levels=levels,
+        keywords={"return_total_data": True},
     )
 
     for i in range(240):
@@ -49,18 +53,42 @@ def test_manybody_full_2(fulltest_client: PortalClient):
         fragments=[[0], [1], [2], [3]],
     )
 
-    sp_spec = {
+    sp_spec_1 = {
         "program": "psi4",
         "driver": "energy",
         "method": "mp2",
-        "basis": "aug-cc-pvdz",
-        "keywords": {"e_convergence": 1e-10, "d_convergence": 1e-10},
+        "basis": "sto-3g",
+        "keywords": {"cc_type": "df", "df_basis_mp2": "def2-qzvpp-ri"},
+    }
+
+    sp_spec_2 = {
+        "program": "psi4",
+        "driver": "energy",
+        "method": "b3lyp",
+        "basis": "sto-3g",
+        "keywords": {"cc_type": "df", "df_basis_mp2": "def2-qzvpp-ri"},
+    }
+
+    sp_spec_3 = {
+        "program": "psi4",
+        "driver": "energy",
+        "method": "hf",
+        "basis": "sto-3g",
+        "keywords": {"cc_type": "df", "df_basis_mp2": "def2-qzvpp-ri"},
     }
 
-    mb_keywords = {"max_nbody": 2, "bsse_correction": "cp"}
+    levels = {
+        1: sp_spec_1,
+        2: sp_spec_2,
+        "supersystem": sp_spec_3,
+    }
 
     meta, ids = fulltest_client.add_manybodys(
-        initial_molecules=[molecule], program="manybody", singlepoint_specification=sp_spec, keywords=mb_keywords
+        initial_molecules=[molecule],
+        program="qcmanybody",
+        bsse_correction=["nocp", "cp", "vmfc"],
+        levels=levels,
+        keywords={"return_total_data": True},
     )
 
     for i in range(240):

qcfractal/pyproject.toml

@@ -35,6 +35,7 @@ dependencies = [
 [project.optional-dependencies]
 services = [
     "torsiondrive",
+    "qcmanybody",
 ]
 geoip = [
     "geoip2"

qcfractal/qcfractal/alembic/versions/2024-06-11-a5a701dc344d_delete_old_manybody.py

@@ -0,0 +1,49 @@
+"""delete old manybody
+
+Revision ID: a5a701dc344d
+Revises: 73b4838a6839
+Create Date: 2024-06-11 15:51:11.380308
+
+"""
+
+from alembic import op
+import sqlalchemy as sa
+
+
+# revision identifiers, used by Alembic.
+revision = "a5a701dc344d"
+down_revision = "73b4838a6839"
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+
+    conn = op.get_bind()
+
+    res = conn.execute(sa.text("SELECT count(*) FROM base_record WHERE record_type = 'manybody'"))
+    count = res.fetchone()[0]
+    if count != 0:
+        raise ValueError("Will not delete old manybody tables with existing data")
+
+    res = conn.execute(sa.text("SELECT count(*) FROM manybody_record"))
+    count = res.fetchone()[0]
+    if count != 0:
+        raise ValueError("Will not delete old manybody tables with existing data")
+
+    res = conn.execute(sa.text("SELECT count(*) FROM manybody_dataset"))
+    count = res.fetchone()[0]
+    if count != 0:
+        raise ValueError("Will not delete old manybody tables with existing data")
+
+    op.execute(sa.text("DROP TABLE manybody_cluster CASCADE"))
+    op.execute(sa.text("DROP TABLE manybody_dataset CASCADE"))
+    op.execute(sa.text("DROP TABLE manybody_dataset_entry CASCADE"))
+    op.execute(sa.text("DROP TABLE manybody_dataset_record CASCADE"))
+    op.execute(sa.text("DROP TABLE manybody_dataset_specification CASCADE"))
+    op.execute(sa.text("DROP TABLE manybody_record CASCADE"))
+    op.execute(sa.text("DROP TABLE manybody_specification CASCADE"))
+
+
+def downgrade():
+    raise NotImplementedError("Downgrade not supported")

qcfractal/qcfractal/alembic/versions/2024-06-11-fd95035b773b_new_manybody_code.py

@@ -0,0 +1,215 @@
+"""new manybody code
+
+Revision ID: fd95035b773b
+Revises: a5a701dc344d
+Create Date: 2024-06-11 16:29:38.468745
+
+"""
+
+import sqlalchemy as sa
+from alembic import op
+from sqlalchemy.dialects import postgresql
+
+# revision identifiers, used by Alembic.
+revision = "fd95035b773b"
+down_revision = "a5a701dc344d"
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.create_table(
+        "manybody_specification",
+        sa.Column("id", sa.Integer(), nullable=False),
+        sa.Column("program", sa.String(), nullable=False),
+        sa.Column("bsse_correction", postgresql.ARRAY(sa.String()), nullable=False),
+        sa.Column("keywords", postgresql.JSONB(astext_type=sa.Text()), nullable=False),
+        sa.Column("protocols", postgresql.JSONB(astext_type=sa.Text()), nullable=False),
+        sa.Column("specification_hash", sa.String(), nullable=False),
+        sa.CheckConstraint("program = LOWER(program)", name="ck_manybody_specification_program_lower"),
+        sa.PrimaryKeyConstraint("id"),
+    )
+    op.create_index("ix_manybody_specification_program", "manybody_specification", ["program"], unique=False)
+    op.create_table(
+        "manybody_specification_levels",
+        sa.Column("id", sa.Integer(), nullable=False),
+        sa.Column("manybody_specification_id", sa.Integer(), nullable=False),
+        sa.Column("level", sa.Integer(), nullable=False),
+        sa.Column("singlepoint_specification_id", sa.Integer(), nullable=False),
+        sa.ForeignKeyConstraint(
+            ["manybody_specification_id"],
+            ["manybody_specification.id"],
+        ),
+        sa.ForeignKeyConstraint(
+            ["singlepoint_specification_id"],
+            ["qc_specification.id"],
+        ),
+        sa.PrimaryKeyConstraint("id"),
+        sa.UniqueConstraint("manybody_specification_id", "level", name="ux_manybody_specification_levels_unique"),
+    )
+    op.create_index(
+        "ix_manybody_specifications_levels_manybody_specification_id",
+        "manybody_specification_levels",
+        ["manybody_specification_id"],
+        unique=False,
+    )
+    op.create_index(
+        "ix_manybody_specifications_levels_singlepoint_specification_id",
+        "manybody_specification_levels",
+        ["singlepoint_specification_id"],
+        unique=False,
+    )
+    op.create_table(
+        "manybody_dataset",
+        sa.Column("id", sa.Integer(), nullable=False),
+        sa.ForeignKeyConstraint(["id"], ["base_dataset.id"], ondelete="cascade"),
+        sa.PrimaryKeyConstraint("id"),
+    )
+    op.create_table(
+        "manybody_record",
+        sa.Column("id", sa.Integer(), nullable=False),
+        sa.Column("initial_molecule_id", sa.Integer(), nullable=False),
+        sa.Column("specification_id", sa.Integer(), nullable=False),
+        sa.ForeignKeyConstraint(["id"], ["base_record.id"], ondelete="cascade"),
+        sa.ForeignKeyConstraint(
+            ["initial_molecule_id"],
+            ["molecule.id"],
+        ),
+        sa.ForeignKeyConstraint(
+            ["specification_id"],
+            ["manybody_specification.id"],
+        ),
+        sa.PrimaryKeyConstraint("id"),
+    )
+    op.create_table(
+        "manybody_cluster",
+        sa.Column("id", sa.Integer(), nullable=False),
+        sa.Column("manybody_id", sa.Integer(), nullable=True),
+        sa.Column("molecule_id", sa.Integer(), nullable=False),
+        sa.Column("mc_level", sa.String(), nullable=False),
+        sa.Column("fragments", postgresql.ARRAY(sa.Integer()), nullable=False),
+        sa.Column("basis", postgresql.ARRAY(sa.Integer()), nullable=False),
+        sa.Column("singlepoint_id", sa.Integer(), nullable=True),
+        sa.CheckConstraint("array_length(basis, 1) > 0", name="ck_manybody_cluster_basis"),
+        sa.CheckConstraint("array_length(fragments, 1) > 0", name="ck_manybody_cluster_fragments"),
+        sa.ForeignKeyConstraint(["manybody_id"], ["manybody_record.id"], ondelete="cascade"),
+        sa.ForeignKeyConstraint(
+            ["molecule_id"],
+            ["molecule.id"],
+        ),
+        sa.ForeignKeyConstraint(
+            ["singlepoint_id"],
+            ["singlepoint_record.id"],
+        ),
+        sa.PrimaryKeyConstraint("id"),
+        sa.UniqueConstraint("manybody_id", "mc_level", "fragments", "basis", name="ux_manybody_cluster_unique"),
+    )
+    op.create_index("ix_manybody_cluster_molecule_id", "manybody_cluster", ["molecule_id"], unique=False)
+    op.create_index("ix_manybody_cluster_singlepoint_id", "manybody_cluster", ["singlepoint_id"], unique=False)
+    op.create_table(
+        "manybody_dataset_entry",
+        sa.Column("dataset_id", sa.Integer(), nullable=False),
+        sa.Column("name", sa.String(), nullable=False),
+        sa.Column("comment", sa.String(), nullable=True),
+        sa.Column("initial_molecule_id", sa.Integer(), nullable=False),
+        sa.Column("additional_singlepoint_keywords", postgresql.JSONB(astext_type=sa.Text()), nullable=False),
+        sa.Column("attributes", postgresql.JSONB(astext_type=sa.Text()), nullable=False),
+        sa.ForeignKeyConstraint(["dataset_id"], ["manybody_dataset.id"], ondelete="cascade"),
+        sa.ForeignKeyConstraint(
+            ["initial_molecule_id"],
+            ["molecule.id"],
+        ),
+        sa.PrimaryKeyConstraint("dataset_id", "name"),
+    )
+    op.create_index("ix_manybody_dataset_entry_dataset_id", "manybody_dataset_entry", ["dataset_id"], unique=False)
+    op.create_index(
+        "ix_manybody_dataset_entry_initial_molecule_id", "manybody_dataset_entry", ["initial_molecule_id"], unique=False
+    )
+    op.create_index("ix_manybody_dataset_entry_name", "manybody_dataset_entry", ["name"], unique=False)
+    op.create_table(
+        "manybody_dataset_specification",
+        sa.Column("dataset_id", sa.Integer(), nullable=False),
+        sa.Column("name", sa.String(), nullable=False),
+        sa.Column("description", sa.String(), nullable=True),
+        sa.Column("specification_id", sa.Integer(), nullable=False),
+        sa.ForeignKeyConstraint(["dataset_id"], ["manybody_dataset.id"], ondelete="cascade"),
+        sa.ForeignKeyConstraint(
+            ["specification_id"],
+            ["manybody_specification.id"],
+        ),
+        sa.PrimaryKeyConstraint("dataset_id", "name"),
+    )
+    op.create_index(
+        "ix_manybody_dataset_specification_dataset_id", "manybody_dataset_specification", ["dataset_id"], unique=False
+    )
+    op.create_index("ix_manybody_dataset_specification_name", "manybody_dataset_specification", ["name"], unique=False)
+    op.create_index(
+        "ix_manybody_dataset_specification_specification_id",
+        "manybody_dataset_specification",
+        ["specification_id"],
+        unique=False,
+    )
+    op.create_table(
+        "manybody_dataset_record",
+        sa.Column("dataset_id", sa.Integer(), nullable=False),
+        sa.Column("entry_name", sa.String(), nullable=False),
+        sa.Column("specification_name", sa.String(), nullable=False),
+        sa.Column("record_id", sa.Integer(), nullable=False),
+        sa.ForeignKeyConstraint(
+            ["dataset_id", "entry_name"],
+            ["manybody_dataset_entry.dataset_id", "manybody_dataset_entry.name"],
+            onupdate="cascade",
+            ondelete="cascade",
+        ),
+        sa.ForeignKeyConstraint(
+            ["dataset_id", "specification_name"],
+            ["manybody_dataset_specification.dataset_id", "manybody_dataset_specification.name"],
+            onupdate="cascade",
+            ondelete="cascade",
+        ),
+        sa.ForeignKeyConstraint(["dataset_id"], ["manybody_dataset.id"], ondelete="cascade"),
+        sa.ForeignKeyConstraint(
+            ["record_id"],
+            ["manybody_record.id"],
+        ),
+        sa.PrimaryKeyConstraint("dataset_id", "entry_name", "specification_name"),
+    )
+    op.create_index("ix_manybody_dataset_record_record_id", "manybody_dataset_record", ["record_id"], unique=False)
+
+    op.execute(
+        """CREATE TRIGGER qca_manybody_record_delete_base_tr
+                  AFTER DELETE ON public.manybody_record
+                  FOR EACH ROW EXECUTE FUNCTION qca_base_record_delete();"""
+    )
+
+    # ### end Alembic commands ###
+
+
+def downgrade():
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.drop_index("ix_manybody_dataset_record_record_id", table_name="manybody_dataset_record")
+    op.drop_table("manybody_dataset_record")
+    op.drop_index("ix_manybody_dataset_specification_specification_id", table_name="manybody_dataset_specification")
+    op.drop_index("ix_manybody_dataset_specification_name", table_name="manybody_dataset_specification")
+    op.drop_index("ix_manybody_dataset_specification_dataset_id", table_name="manybody_dataset_specification")
+    op.drop_table("manybody_dataset_specification")
+    op.drop_index("ix_manybody_dataset_entry_name", table_name="manybody_dataset_entry")
+    op.drop_index("ix_manybody_dataset_entry_initial_molecule_id", table_name="manybody_dataset_entry")
+    op.drop_index("ix_manybody_dataset_entry_dataset_id", table_name="manybody_dataset_entry")
+    op.drop_table("manybody_dataset_entry")
+    op.drop_index("ix_manybody_cluster_singlepoint_id", table_name="manybody_cluster")
+    op.drop_index("ix_manybody_cluster_molecule_id", table_name="manybody_cluster")
+    op.drop_table("manybody_cluster")
+    op.drop_table("manybody_record")
+    op.drop_table("manybody_dataset")
+    op.drop_index(
+        "ix_manybody_specifications_levels_singlepoint_specification_id", table_name="manybody_specification_levels"
+    )
+    op.drop_index(
+        "ix_manybody_specifications_levels_manybody_specification_id", table_name="manybody_specification_levels"
+    )
+    op.drop_table("manybody_specification_levels")
+    op.drop_index("ix_manybody_specification_program", table_name="manybody_specification")
+    op.drop_table("manybody_specification")
+    # ### end Alembic commands ###

qcfractal/qcfractal/alembic/versions/2025-01-14-d5988aa750ae_merge_manybody_branch_with_changes.py

@@ -0,0 +1,21 @@
+"""merge manybody branch with changes
+
+Revision ID: d5988aa750ae
+Revises: 3690c677f8d1, fd95035b773b
+Create Date: 2025-01-14 10:49:27.547435
+
+"""
+
+# revision identifiers, used by Alembic.
+revision = "d5988aa750ae"
+down_revision = ("3690c677f8d1", "fd95035b773b")
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+    pass
+
+
+def downgrade():
+    pass

qcfractal/qcfractal/components/manybody/dataset_db_models.py

@@ -19,7 +19,7 @@ class ManybodyDatasetEntryORM(BaseORM):
     comment = Column(String)
 
     initial_molecule_id = Column(Integer, ForeignKey(MoleculeORM.id), nullable=False)
-    additional_keywords = Column(JSONB, nullable=False)
+    additional_singlepoint_keywords = Column(JSONB, nullable=False)
     attributes = Column(JSONB, nullable=False)
 
     initial_molecule = relationship(MoleculeORM, lazy="joined")