Merge branch 'main' of github.com:AI2MS/pyAKI

Author: aegis301

README.md

@@ -5,6 +5,7 @@
 Python package to detect AKI within time series data.
 
 The goal of this package is to establish well tested, comprehensive functions for the detection of Acute Kidney Injury (AKI) in time series data, according to the Kidney Disease Improving Global Outcomes (KDIGO) Criteria, established in 2012 [^kdigo].
+![](img/kdigo_criteria.png)
 
 ## Installation
 
@@ -30,6 +31,4 @@ process_stay(stay_id: int,
 PYTHONPATH=".:${PYTHONPATH}" python -m unittest discover
 ```
 
-![kdigo_criteria](img/kdigo_criteria.png)
-
-[^kdigo]: Khwaja A. KDIGO clinical practice guidelines for acute kidney injury. Nephron Clin Pract. 2012;120(4):c179-84. doi: 10.1159/000339789. Epub 2012 Aug 7. PMID: 22890468.
+[^kdigo]: Improving Global Outcomes (KDIGO) Acute Kidney Injury Work Group. KDIGO Clinical Practice Guideline for Acute Kidney Injury. Kidney inter., Suppl. 2012; 2: 1–138.

pyAKI/kdigo.py

@@ -1,3 +1,5 @@
+import logging
+
 from typing import Optional
 
 import pandas as pd
@@ -20,6 +22,8 @@
 
 from pyAKI.utils import Dataset
 
+logger = logging.getLogger(__name__)
+
 
 class Analyser:
     """
@@ -89,9 +93,12 @@ def __init__(
         self.validate_data(data)
 
         # apply preprocessors to the input data
+        logger.info("Start preprocessing")
         for preprocessor in preprocessors:
             data = preprocessor.process(data)
 
+        logger.info("Finish preprocessing")
+
         self._data: list[Dataset] = data
         self._probes: list[Probe] = probes
         self._stay_identifier: str = stay_identifier
@@ -111,6 +118,8 @@ def process_stays(self) -> pd.DataFrame:
         Returns:
             pd.DataFrame: The analysis results for all stays.
         """
+        logger.info("Start probing")
+
         (_, df), *datasets = self._data
         stay_ids: pd.Index = df.index.get_level_values("stay_id").unique()
         for _, df in datasets:
@@ -119,7 +128,8 @@ def process_stays(self) -> pd.DataFrame:
         data: pd.DataFrame = self.process_stay(stay_ids.values[0])
         for stay_id in stay_ids.values[1:]:
             data = pd.concat([data, self.process_stay(stay_id)])
-        # data = data.drop(columns=["stay_id"])  # remove additional stay_id column
+
+        logger.info("Finish probing")
         return data
 
     def process_stay(self, stay_id: str) -> pd.DataFrame:
@@ -135,6 +145,7 @@ def process_stay(self, stay_id: str) -> pd.DataFrame:
         Returns:
             pd.DataFrame: The analysis results for the specific stay.
         """
+        logger.debug("Processing stay with id: %s", stay_id)
 
         datasets: list[Dataset] = [
             Dataset(dtype, data.loc[stay_id]) for dtype, data in self._data  # type: ignore

pyAKI/utils.py

@@ -1,10 +1,14 @@
+import logging
+
 from typing import NamedTuple, cast
 from enum import StrEnum, auto
 from functools import wraps
 
 import numpy as np
 import pandas as pd
 
+logger = logging.getLogger(__name__)
+
 
 class DatasetType(StrEnum):
     """Enumeration class representing different types of datasets."""
@@ -89,6 +93,10 @@ def wrapper(self, datasets: list[Dataset], *args, **kwargs) -> list[Dataset]:
             }
             # check if all datasets are mapped, otherwise return the original datasets
             if len(in_mapping) != len(_mapping):
+                logger.warning(
+                    "Skip %s because one or more datasets are missing to probe",
+                    self.__class__.__name__,
+                )
                 return datasets
 
             # call the wrapped function with the converted DataFrames