pycomorb 0.5.2__tar.gz → 0.6__tar.gz

{pycomorb-0.5.2/src/pycomorb.egg-info → pycomorb-0.6}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pycomorb
-Version: 0.5.2
+Version: 0.6
 Summary: Comorbidity and frailty score calculation utilities (Charlson, Elixhauser, Gagne, HFRS) using polars
 Author-email: Finn Fassbender <finn.fassbender@charite.de>
 License-Expression: MIT

{pycomorb-0.5.2 → pycomorb-0.6}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "pycomorb"
-version = "0.5.2"
+version = "0.6"
 authors = [
   { name="Finn Fassbender", email="finn.fassbender@charite.de" },
 ]
@@ -23,4 +23,6 @@ requires = ["setuptools >= 77.0.3"]
 build-backend = "setuptools.build_meta"
 
 [tool.setuptools.package-data]
-"pycomorb.common" = ["*.csv"]
+"pycomorb.common" = ["*.csv"]
+"pycomorb.modification_DE" = ["*.csv"]
+"pycomorb.modification_US" = ["*.csv"]

{pycomorb-0.5.2 → pycomorb-0.6}/src/pycomorb/CharlsonComorbidityIndex.py

@@ -28,7 +28,6 @@
 
 from pathlib import Path
 
-import pandas as pd
 import polars as pl
 
 # Import the generalized function
@@ -63,11 +62,6 @@ def CharlsonComorbidityIndex(
         - DataFrame with category indicators if return_categories is True, else None.
     """
 
-    # Check if input is pandas DataFrame and convert to polars
-    is_pandas = pd and isinstance(df, pd.DataFrame)
-    if is_pandas:
-        df = pl.from_pandas(df)
-
     # Change ICD to ICD-9 for Deyo, D'Hoore and Romano
     if icd_version == "icd10" and implementation in [
         "deyo",
@@ -210,9 +204,6 @@ def CharlsonComorbidityIndex(
     if not return_categories:
         df_charlson = df_charlson.drop("Age Score")
 
-    if is_pandas:
-        return df_charlson.to_pandas()
-
     return df_charlson
 
 

{pycomorb-0.5.2 → pycomorb-0.6}/src/pycomorb/CustomComorbidityIndex.py

@@ -1,7 +1,6 @@
 from pathlib import Path
 from typing import List, Optional, Tuple, Union
 
-import pandas as pd
 import polars as pl
 
 
@@ -37,11 +36,6 @@ def CustomComorbidityIndex(
         - DataFrame with category indicators if return_categories is True, else None.
     """
 
-    # Check if input is pandas DataFrame and convert to polars
-    is_pandas = pd and isinstance(df, pd.DataFrame)
-    if is_pandas:
-        df = pl.from_pandas(df)
-
     # Input validation
     assert icd_version in [
         "icd9",
@@ -226,7 +220,4 @@ def CustomComorbidityIndex(
         {score_col_name: int}
     )
 
-    if is_pandas:
-        comorbidity_df = comorbidity_df.to_pandas()
-
     return comorbidity_df

{pycomorb-0.5.2 → pycomorb-0.6}/src/pycomorb/ElixhauserComorbidityIndex.py

@@ -18,7 +18,6 @@
 
 from pathlib import Path
 
-import pandas as pd
 import polars as pl
 
 # Import the generalized function
@@ -53,11 +52,6 @@ def ElixhauserComorbidityIndex(
         - DataFrame with category indicators if return_categories is True, else None.
     """
 
-    # Check if input is pandas DataFrame and convert to polars
-    is_pandas = pd and isinstance(df, pd.DataFrame)
-    if is_pandas:
-        df = pl.from_pandas(df)
-
     # Change ICD to ICD-9 for original Elixhauser and AHRQ ICD-9
     if icd_version == "icd10" and implementation in [
         "elixhauser",
@@ -147,7 +141,4 @@ def ElixhauserComorbidityIndex(
         return_categories=return_categories,
     )
 
-    if is_pandas:
-        return df_elixhauser.to_pandas()
-
     return df_elixhauser

{pycomorb-0.5.2 → pycomorb-0.6}/src/pycomorb/GagneComorbidityIndex.py

@@ -22,7 +22,6 @@
 
 from pathlib import Path
 
-import pandas as pd
 import polars as pl
 
 # Import the generalized function
@@ -54,11 +53,6 @@ def GagneComorbidityIndex(
         - DataFrame with category indicators if return_categories is True, else None.
     """
 
-    # Check if input is pandas DataFrame and convert to polars
-    is_pandas = pd and isinstance(df, pd.DataFrame)
-    if is_pandas:
-        df = pl.from_pandas(df)
-
     assert (
         id_col in df.columns
     ), f"Column '{id_col}' (ID) must be present in input DataFrame."
@@ -90,7 +84,4 @@ def GagneComorbidityIndex(
         return_categories=return_categories,
     )
 
-    if is_pandas:
-        return df_gagne.to_pandas()
-
     return df_gagne

pycomorb-0.6/src/pycomorb/ICDModifications.py

@@ -0,0 +1,128 @@
+# Reference for ICD-10-CM modifications:
+# CMS: https://www.cms.gov/medicare/coding-billing/icd-10-codes
+
+# Reference for ICD-10-GM modifications:
+# BfArM: https://www.bfarm.de/DE/Kodiersysteme/Services/Downloads/_node.html
+
+from pathlib import Path
+
+import polars as pl
+
+
+def get_icdmodification(
+    data: pl.DataFrame,
+    transfer_file_path: str,
+    code_col="code",
+    year_col="year",
+    target_year=2024,
+) -> pl.DataFrame:
+    """
+    Map ICD codes to the target year using official transfer files.
+
+    Args:
+        data (pl.DataFrame): Input DataFrame with at least columns [code_col, year_col].
+        transfer_file_path (str): Path to the CSV file containing the transfer mappings.
+        code_col (str): Name of the column containing ICD codes. Default: "code".
+        year_col (str): Name of the column containing the year of the ICD code. Default: "year".
+        target_year (int): Target year for ICD codes (e.g., 2024). Default: 2024.
+
+    Returns:
+        pl.DataFrame: DataFrame with original columns plus a new column 'icdmod_{target_year}' with mapped codes.
+    """
+
+    assert (
+        code_col in data.columns
+    ), f"Column '{code_col}' (ICD code) must be present in input DataFrame."
+    assert (
+        data.height
+        == data.filter(pl.col(code_col).str.contains(r"^[A-Z]")).height
+    ), f"All values in column '{code_col}' must start with an uppercase letter (A-Z)."
+    assert (
+        year_col in data.columns
+    ), f"Column '{year_col}' (year) must be present in input DataFrame."
+
+    # Unpivot the data to long format
+    transfer_data = (
+        pl.read_csv(transfer_file_path, null_values=["", " "], separator=",")
+        .unpivot(
+            index=str(target_year),
+            variable_name=year_col,
+            value_name=code_col,
+        )
+        .cast({year_col: int})
+        .rename({str(target_year): f"icd_{target_year}"})
+    )
+
+    return (
+        data.join(transfer_data, on=[code_col, year_col], how="left")
+        # Prefer mapped code, but fall back to original if no mapping available
+        .with_columns(
+            pl.coalesce(pl.col(f"icd_{target_year}"), pl.col(code_col)).alias(
+                f"icd_{target_year}"
+            )
+        )
+    )
+
+
+def get_icd10gm(
+    data: pl.DataFrame, code_col="code", year_col="year", target_year=2024
+) -> pl.DataFrame:
+    """
+    Map ICD-10-GM codes to the target year using official transfer files.
+
+    Args:
+        data (pl.DataFrame): Input DataFrame with at least columns [code_col, year_col].
+        code_col (str): Name of the column containing ICD codes. Default: "code".
+        year_col (str): Name of the column containing the year of the ICD code. Default: "year".
+        target_year (int): Target year for ICD-10-GM codes (e.g., 2024). Default: 2024.
+
+    Returns:
+        pl.DataFrame: DataFrame with original columns plus a new column 'icd10gm_{target_year}' with mapped codes.
+    """
+
+    assert target_year in range(
+        2004, 2026
+    ), "Target year must be between 2004 and 2025."
+
+    # Load all transfer files into a dictionary of DataFrames
+    transfer_file_path = Path(__file__).parent / "modification_DE/icd10gm.csv" # fmt: skip
+
+    return get_icdmodification(
+        data=data,
+        transfer_file_path=transfer_file_path,
+        code_col=code_col,
+        year_col=year_col,
+        target_year=target_year,
+    )
+
+
+def get_icd10cm(
+    data: pl.DataFrame, code_col="code", year_col="year", target_year=2024
+) -> pl.DataFrame:
+    """
+    Map ICD-10-CM codes to the target year using official transfer files.
+
+    Args:
+        data (pl.DataFrame): Input DataFrame with at least columns [code_col, year_col].
+        code_col (str): Name of the column containing ICD codes. Default: "code".
+        year_col (str): Name of the column containing the year of the ICD code. Default: "year".
+        target_year (int): Target year for ICD-10-CM codes (e.g., 2024). Default: 2024.
+
+    Returns:
+        pl.DataFrame: DataFrame with original columns plus a new column 'icd10gm_{target_year}' with mapped codes.
+    """
+
+    assert target_year in range(
+        2015, 2026
+    ), "Target year must be between 2004 and 2025."
+
+    # Load all transfer files into a dictionary of DataFrames
+    transfer_file_path = Path(__file__).parent / "modification_US/icd10cm.csv" # fmt: skip
+
+    return get_icdmodification(
+        data=data,
+        transfer_file_path=transfer_file_path,
+        code_col=code_col,
+        year_col=year_col,
+        target_year=target_year,
+    )

{pycomorb-0.5.2 → pycomorb-0.6}/src/pycomorb/__init__.py

@@ -2,14 +2,17 @@
 
 __author__ = """Finn Fassbender"""
 __email__ = "finn.fassbender@charite.de"
-__version__ = "0.5.2"
+__version__ = "0.6"
+
+import pandas as pd
+import polars as pl
 
 from .CharlsonComorbidityIndex import CharlsonComorbidityIndex
+from .CustomComorbidityIndex import CustomComorbidityIndex
 from .ElixhauserComorbidityIndex import ElixhauserComorbidityIndex
 from .GagneComorbidityIndex import GagneComorbidityIndex
-from .CustomComorbidityIndex import CustomComorbidityIndex
 from .HospitalFrailtyRiskScore import HospitalFrailtyRiskScore
-from .ICDModifications import get_icd10gm
+from .ICDModifications import get_icd10cm, get_icd10gm
 
 
 def comorbidity(
@@ -31,6 +34,7 @@ def comorbidity(
     score_col_name: str = "Custom Comorbidity Score",
     mutual_exclusion_rules: list[tuple[str, str]] = None,
     return_categories: bool = False,
+    fix_dot_in_icd_code: bool = False,
 ):
     """
     Unified wrapper to calculate a comorbidity or frailty score.
@@ -42,7 +46,7 @@ def comorbidity(
         code_col (str): Name of the column containing ICD codes. Default: "code".
         age_col (str): Name of the column containing patient ages (Charlson only). Default: "age".
         year_col (str): Name of the column containing the year of the ICD code (for ICD history modifications). Default: "year".
-        icd_version (str): ICD version. Default: "icd10".
+        icd_version (str): ICD version. Must be one of: 'icd9', 'icd10', or 'icd9_10'. Default: "icd10".
         icd_version_col (str, optional): Name of the column with ICD version for 'icd9_10'. Default: None.
         icd_modification (str, optional): ICD modification to apply ('icd10gm'). Default: None.
         icd_modification_target_year (int): Target year for ICD modification (if applicable). Default: 2004.
@@ -54,23 +58,57 @@ def comorbidity(
         score_col_name (str): Name for the calculated score column (Custom only).
         mutual_exclusion_rules (list of tuple, optional): List of mutually exclusive category pairs (Custom only).
         return_categories (bool): Whether to return category indicators.
+        fix_dot_in_icd_code (bool): Whether to remove dots from ICD codes before processing. Default: False.
 
     Returns:
         - DataFrame with [id_col, score].
         - DataFrame with category indicators if return_categories is True, else None.
     """
+
+    # Check if input is pandas DataFrame and convert to polars
+    is_pandas = pd and isinstance(df, pd.DataFrame)
+    if is_pandas:
+        df = pl.from_pandas(df)
+
+    # remove dots from ICD codes if requested
+    if fix_dot_in_icd_code:
+        df = df.with_columns(
+            pl.col(code_col).str.replace(".", "", literal=True)
+        )
+
+    # check that no dots are present in ICD codes (Polars-native)
+    contains_dot = df.select(
+        pl.col(code_col).str.contains(r"\.").any().alias("contains_dot")
+    ).to_dicts()[0]["contains_dot"]
+    assert (
+        not contains_dot
+    ), f"All values in column '{code_col}' must not contain dots ('.'). Consider setting fix_dot_in_icd_code=True."
+    
+    # strip whitespace from ICD codes, and make uppercase
+    df = df.with_columns(
+        pl.col(code_col).str.strip_chars().str.to_uppercase()
+    )
+
     # apply ICD modification if requested
-    if icd_modification is not None:
-        if icd_modification.lower() == "icd10gm":
+    if icd_modification is not None and "10" in icd_version:
+        if icd_modification.lower() == "icd10cm":
+            df = get_icd10cm(
+                data=df,
+                code_col=code_col,
+                year_col=year_col,
+                target_year=icd_modification_target_year,
+            )
+            code_col = f"icd_{icd_modification_target_year}"
+        elif icd_modification.lower() == "icd10gm":
             df = get_icd10gm(
                 data=df,
                 code_col=code_col,
                 year_col=year_col,
                 target_year=icd_modification_target_year,
             )
-            code_col = f"icd10gm_{icd_modification_target_year}"
+            code_col = f"icd_{icd_modification_target_year}"
         else:
-            raise ValueError(f"Unknown icd_modification: '{icd_modification}'. Currently, only 'icd10gm' is supported.") # fmt: skip
+            raise ValueError(f"Unknown icd_modification: '{icd_modification}'. Currently, only 'icd10gm' and 'icd10cm' are supported.") # fmt: skip
 
     # calculate scores
     score = score.lower()
@@ -82,7 +120,7 @@ def comorbidity(
     ):
         if age_col not in df.columns:
             raise ValueError(f"Column '{age_col}' (age) must be present in input DataFrame for Charlson calculation.") # fmt: skip
-        return CharlsonComorbidityIndex(
+        return_df = CharlsonComorbidityIndex(
             df=df,
             id_col=id_col,
             code_col=code_col,
@@ -98,7 +136,7 @@ def comorbidity(
         "elixhausercomorbidityindex",
         "elixhauser_comorbidity_index",
     ):
-        return ElixhauserComorbidityIndex(
+        return_df = ElixhauserComorbidityIndex(
             df=df,
             id_col=id_col,
             code_col=code_col,
@@ -117,7 +155,7 @@ def comorbidity(
         "combinedcomorbidityindex",
         "combined_comorbidity_index",
     ):
-        return GagneComorbidityIndex(
+        return_df = GagneComorbidityIndex(
             df=df,
             id_col=id_col,
             code_col=code_col,
@@ -131,7 +169,7 @@ def comorbidity(
         "hospitalfrailtyriskscore",
         "hospital_frailty_risk_score",
     ):
-        return HospitalFrailtyRiskScore(
+        return_df = HospitalFrailtyRiskScore(
             df=df,
             id_col=id_col,
             code_col=code_col,
@@ -143,7 +181,7 @@ def comorbidity(
         "customcomorbidityindex",
         "custom_comorbidity_index",
     ):
-        return CustomComorbidityIndex(
+        return_df = CustomComorbidityIndex(
             df=df,
             id_col=id_col,
             code_col=code_col,
@@ -160,3 +198,8 @@ def comorbidity(
             f"Unknown score: '{score}'. Must be one of: "
             "'charlson', 'elixhauser', 'gagne', 'hfrs' or 'custom'."
         )
+
+    if is_pandas:
+        return return_df.to_pandas()
+
+    return return_df

{pycomorb-0.5.2 → pycomorb-0.6}/src/pycomorb/common/CHARLSON_QUAN.csv

@@ -1,7 +1,7 @@
 index,category,icd9_codes,icd10_codes,weights
 0,Age,XXXX|XXXX,YYYY|YYYY,1
 1,Myocardial infarction,410|412,I21|I22|I252,1
-2,Congestive heart failure,428|4254|4259|39891|40201|40211|40291|40401|40403|40411|40413|40491|40493,I43|I50|I099|I110|I130|I132|I255|I420|I425|I426|I427|I428|I429|P290,1
+2,Congestive heart failure,428|4254|4255|4256|4257|4258|4259|39891|40201|40211|40291|40401|40403|40411|40413|40491|40493,I43|I50|I099|I110|I130|I132|I255|I420|I425|I426|I427|I428|I429|P290,1
 3,Peripheral vascular disease,440|441|0930|4373|4471|5571|5579|V434|4431|4432|4433|4434|4435|4436|4437|4438|4439,I70|I71|I731|I738|I739|I771|I790|I792|K551|K558|K559|Z958|Z959,1
 4,Cerebrovascular disease,430|431|432|433|434|435|436|437|438|36234,G45|G46|I60|I61|I62|I63|I64|I65|I66|I67|I68|I69|H340,1
 5,Dementia,290|2941|3312,F00|F01|F02|F03|G30|F051|G311,1

{pycomorb-0.5.2/src/pycomorb/germanmodification → pycomorb-0.6/src/pycomorb/modification_DE}/germanmodification.py

@@ -3,7 +3,6 @@
 # 3. take only the transfer files (e.g. icd10gm2013syst_umsteiger_2012_2013.txt for 2012->2013) and put them into a separate folder, e.g. data/icd10gm/transfers
 # 4. run this script in that folder
 
-import os
 import polars as pl
 
 files = [