feat: Add uris field for identifiers (#103)

* Add `uris` field for identifiers

* Linting

* update valdation

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* feat: fix typo

---------

Co-authored-by: Matthew Evans <git@ml-evs.science>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Michael Pieler <Michael.Pieler@Gmail.com>

Author: 4 people

pyproject.toml

@@ -30,6 +30,7 @@ dev = [
         "pre-commit",
         "pydantic_yaml",
         "pytest",
+        "pubchempy"
 ]
 
 # [project.optional-dependencies]

src/chemnlp/data_val/model.py

@@ -1,5 +1,7 @@
 from typing import Dict, List, Optional
 
+import pubchempy as pcp
+import requests
 from pydantic import root_validator, validator
 from pydantic_yaml import YamlModel, YamlStrEnum
 
@@ -19,6 +21,8 @@ class Identifier(YamlModel):
     """Identifier information."""
 
     id: str
+
+    """A description of the field"""
     description: Optional[str]
     type: IdentifierEnum
     names: Optional[List[str]]
@@ -43,17 +47,80 @@ class ColumnTypes(YamlStrEnum):
     continuous = "continuous"
     categorical = "categorical"
     ordinal = "ordinal"
+    boolean = "boolean"
 
 
 class Target(YamlModel):
     """Target information."""
 
     id: str
+
+    """A english description of the field"""
     description: str
-    units: str
+
+    """The units of the field. None if unitless."""
+    units: Optional[str]
+
+    """The type of the field. Can be one of `continuous`, `categorical`, `ordinal`, `boolean`."""
     type: ColumnTypes
+
+    """A list of names describing the field.
+
+    Note that this will be used in building the prompts. Some example for prompts:
+
+    - Boolean variables
+
+        - `Is <name> <identifier>?`
+        - ```
+        What molecules in the list are <name>?
+        - <identifier_1>
+        - <identifier_2>
+        - <identifier_3>
+        ```
+
+
+    - Continuous variables
+
+        - `What is <name> of <identifier>?`
+        - ```
+        What is the molecule with largest <name> in the following list?
+        - <identifier_1>
+        - <identifier_2>
+        - <identifier_3>
+        ```
+    """
     names: List[str]
 
+    """A URI or multiple (consitent ) URIs for the field.
+
+    Ideally this would be a link to an entry in an ontrology or controlled
+    vocabulary that can also provide a canonical description for the field.
+    """
+    uris: Optional[List[str]]
+
+    """A PubChem assay IDs or multiple (consistent) PubChem assay IDs.
+
+    Make sure that the first assay ID is the primary assay ID.
+    """
+    pubchem_aids: Optional[List[int]]
+
+    @validator("pubchem_aids")
+    def uris_resolves(cls, values):
+        if values is not None:
+            for uri in values.get("uris"):
+                # perform a request to the URI and check if it resolves
+                response = requests.get(uri)
+                if response.status_code != 200:
+                    raise ValueError(f"URI {uri} does not resolve")
+
+    @validator("pubchem_aids")
+    def pubchem_assay_ids_resolve(cls, values):
+        if values is not None:
+            for aid in values:
+                assays = pcp.get_assays(aid)
+                if len(assays) == 0:
+                    raise ValueError(f"PubChem assay ID {aid} does not resolve")
+
 
 class Template(YamlModel):
     prompt: str
@@ -95,3 +162,12 @@ class Dataset(YamlModel):
     def num_points_must_be_positive(cls, v):
         if v < 0:
             raise ValueError("num_points must be positive")
+
+    @validator("links")
+    def links_must_resolve(cls, v):
+        if v is not None:
+            for link in v:
+                response = requests.get(link.url)
+                if response.status_code != 200:
+                    if not (("acs" in response.text) or ("sage" in response.text)):
+                        raise ValueError(f"Link {link.url} does not resolve")