From 825ff186f2a41c06e7eb9c67ccd4097703c196d4 Mon Sep 17 00:00:00 2001
From: Dawith Lim <lim185@purdue.edu>
Date: Tue, 30 Sep 2025 13:00:36 -0400
Subject: [PATCH] Function descriptions added

---
 pipe/extract.py | 34 +++++++++++++++++++++++++++++++---
 1 file changed, 31 insertions(+), 3 deletions(-)

diff --git a/pipe/extract.py b/pipe/extract.py
index 9415ee9..849246e 100644
--- a/pipe/extract.py
+++ b/pipe/extract.py
@@ -14,11 +14,13 @@
 from pyspark.sql import SparkSession, Row, DataFrame
 import scipy as sp
 
-def extract(spark):
+def extract(spark: SparkSession) -> DataFrame:
     """
     First step of the ETL pipeline. It reads the list of .mat files from
     a CSV list, opens and pulls the spectrogram from each respective file.
 
+    Args:
+        spark (SparkSession): Spark session object.
     """
 
     path = Path("/app/workdir")
@@ -33,13 +35,34 @@ def extract(spark):
 
 def image_pipe(spark: SparkSession, imagepath: Path, namepattern: str,
                stacksize: int) -> np.ndarray:
-    images = np.zeros((stacksize, 800,800))
+    """
+    Loads a stack of images from a path based on the given name pattern.
+
+    Args:
+        imagepath (Path): Path to the image files.
+        namepattern (str): Name pattern for the image files.
+        stacksize (int): Number of images in the stack.
+
+    Returns:
+        images: 3D numpy array of stacked images.
+    """
+
+    images = np.zeros((stacksize, 800, 800))
     for i in range(stacksize):
         images[i,:,:] = cv.imread(imagepath/namepattern.format(i), -1)
 
     return images
 
 class SpectrogramReader:
+    """
+    Class to read spectrograms and metadata from different file formats based
+    on user specified filetype.
+
+    Args:
+        spark (SparkSession): Spark session object.
+        filetype (str): File format type. Supported types are 'hdf5',
+                'shards', and 'matfiles'.
+    """
 
     def __init__(self, spark: SparkSession, filetype: str = "hdf5"):
         self.spark = spark
@@ -86,7 +109,8 @@ def spectrogram_read_matfiles(self, specpath: Path, labels:list,
             labels (list): List of target labels.
 
         Returns:
-            
+            DataFrame: Spark DataFrame containing the spectrograms and
+                    associated metadata.
         """
         spectrograms = []
         row = {}
@@ -127,6 +151,8 @@ def spectrogram_read_hdf5(self, specpath: Path, labels: list,
             specpath (Path): Path to the spectrogram files.
 
         Returns:
+            DataFrame: Spark DataFrame containing the spectrograms and
+                    associated metadata.
         """
         
         metadata = self.metadata_pipe(specpath, labels)
@@ -156,6 +182,8 @@ def spectrogram_read_shards(self, specpath: Path, namepattern: str,
                     spectrogram.
 
         Returns:
+            DataFrame: Spark DataFrame containing the spectrograms and
+                    associated metadata.
         """
 
         return