Merge branch 'main' into text-feature-describer

pyproject.toml

@@ -22,18 +22,18 @@ requires-python = ">=3.9"
 dependencies = [
   "wasabi>=0.9.1,<1.1.2",
   "scikit-learn>=1.1.2,<1.2.3",
-  "wandb>=0.12.0,<0.14.3",
+  "wandb>=0.12.0,<0.15.1",
   "deepchecks>=0.8.0,<0.12.0",
   "dill>=0.3.0,<0.3.7",
   "srsly>=2.4.4,<2.4.6",
-  "SQLAlchemy>=1.4.41, <2.0.10",
+  "SQLAlchemy>=1.4.41, <2.0.11",
   "pyodbc>=4.0.34, <4.0.40",
   "dask>=2022.9.0,<2023.5.0",
   "pydantic>=1.9.0, <1.11.0",
   "psutil>=5.9.1, <6.0.0",
   "pandas>=1.4.0,<2.1.0",
   "catalogue>=2.0.0, <2.1.0",
-  "numpy>=1.23.3,<1.24.2",
+  "numpy>=1.23.3,<1.24.4",
   "transformers>=4.22.2, <5.0.0",
   "pyarrow>=9.0.0,<11.1.0",
   "psycopmlutils>=0.2.4,<0.12.0",
@@ -47,7 +47,7 @@ dev = [
   "pylint",
   "ruff",
   "black>=22.8.0,<23.1.1",
-  "docformatter>=1.5.0, <1.6.1",
+  "docformatter>=1.5.0, <1.6.4",
   "pre-commit==3.2.2",
   "mypy==1.2.0",
   "pydantic"

src/psycop_feature_generation/loaders/example/load_lab_results.py

@@ -0,0 +1,6 @@
+"""Example of loading lab results."""
+
+import psycop_feature_generation.loaders.raw.load_lab_results as lb
+
+if __name__ == "__main__":
+    df = lb.cancelled_standard_lab_results(n_rows=5000)

src/psycop_feature_generation/loaders/raw/load_lab_results.py

@@ -123,7 +123,7 @@ def load_cancelled(
     else:
         npu_where = f"npukode = '{blood_sample_id}'"
 
-    sql = f"SELECT {cols} FROM [fct].{view} {npu_where} AND datotid_sidstesvar IS NOT NULL AND Svar == 'Aflyst' AND (left(Svar,1) == '>' OR left(Svar, 1) == '<')"
+    sql = f"SELECT {cols} FROM [fct].{view} WHERE {npu_where} AND datotid_sidstesvar IS NOT NULL AND Svar = 'Aflyst'"
 
     df = sql_load(
         sql,
@@ -753,3 +753,46 @@ def p_ethanol(
         n_rows=n_rows,
         values_to_load=values_to_load,
     )
+
+
+@data_loaders.register("cancelled_standard_lab_results")
+def cancelled_standard_lab_results(
+    n_rows: int | None = None,
+    values_to_load: str = "cancelled",
+) -> pd.DataFrame:
+    """Cancelled lab results for some standard lab results: HbA1c, triglycerides, HDL, LDL, ALAT, ASAT,
+    lymphocytes, leukocytes, crp, creatinine, egfr, albumine-creatinine-ratio, haemoglobin, natrium, kalium,
+    calcium, trombocytes, d-vitamin, tsh, b12-vitamin"""
+    return blood_sample(
+        blood_sample_id=[
+            "NPU27300",  # hba1c
+            "AAB00093",  # hba1c
+            "NPU04094",  # triglycerides
+            "NPU01567",  # hdl
+            "NPU01568",  # ldl
+            "AAB00101",  # ldl
+            "NPU19651",  # alat
+            "NPU19654",  # asat
+            "NPU02636",  # lymphocytes
+            "NPU02593",  # leukocytes
+            "NPU19748",  # crp
+            "NPU18016",  # creatinine
+            "ASS00355",  # creatinine
+            "ASS00354",  # creatinine
+            "DNK35302",  # egfr
+            "DNK35131",  # egfr
+            "AAB00345",  # egfr
+            "AAB00343",  # egfr
+            "NPU19661",  # albumine_creatinine_ratio
+            "NPU02319",  # haemoglobin
+            "NPU03429",  # natrium
+            "NPU03230",  # kalium
+            "NPU01443",  # calcium
+            "NPU03568",  # trombocytes
+            "NPU10267",  # d-vitamin
+            "NPU03577",  # tsh
+            "NPU01700",  # b12-vitamin
+        ],
+        n_rows=n_rows,
+        values_to_load=values_to_load,
+    )

src/psycop_feature_generation/loaders/raw/load_text.py

@@ -121,7 +121,7 @@ def load_text_sfis(
 
     df = pd.concat(dfs)
     df = df.rename(
-        {"datotid_senest_aendret_i_sfien": "timestamp", "fritekst": "text"},
+        {"datotid_senest_aendret_i_sfien": "timestamp", "fritekst": "value"},
         axis=1,
     )
     return df