varsha-preprocess 0.1.1__tar.gz

varsha_preprocess-0.1.1/My_Preprocess1/Column_Transformer.py

@@ -0,0 +1,58 @@
+from .encoders.MyLabelEncoder import MyLabelEncoder
+from .encoders.MyOneHotEncoder import MyOneHotEncoder
+from .encoders.MyOrdinalEncoder import MyOrdinalEncoder
+from .my_pipeline import MyPipeline
+
+import numpy as np
+import pandas as pd
+
+class MyColumnTransformer:
+
+    def __init__(self, transformers):
+        self.transformers = transformers
+
+    def fit_transform(self, df):
+        outputs = []
+
+        for name, encoder, col in self.transformers:
+            
+            # it will extract column data
+            data = df[col]
+            # data = np.array(data).reshape(-1, 1)
+            transformed = encoder.fit_transform(data)
+
+            outputs.append(transformed)
+
+        # this will combine all outputs
+        return np.concatenate(outputs, axis=1)
+
+
+# df = pd.read_csv("Practice/SalaryData.csv")
+
+# ct = MyColumnTransformer([
+#     # ("label", MyLabelEncoder(), "cough"),
+#     ("ordinal", MyOneHotEncoder(), "Education Level")
+#     # ("onehot", MyOneHotEncoder(), "city")
+# ])
+
+
+# result = ct.fit_transform(df)
+# print(result) 
+
+
+# ct = MyColumnTransformer([
+
+#     ("education_pipe",
+#      MyPipeline([
+#          ("ordinal", MyOrdinalEncoder())
+#      ]),
+#      "Education Level"
+#     ),
+
+#     ("gender_pipe",
+#      MyPipeline([
+#          ("onehot", MyOneHotEncoder())
+#      ]),
+#      "Gender"
+#     )])
+

varsha_preprocess-0.1.1/My_Preprocess1/__init__.py

@@ -0,0 +1,9 @@
+from .encoders.MyLabelEncoder import MyLabelEncoder
+from .encoders.MyOneHotEncoder import MyOneHotEncoder
+from .encoders.MyOrdinalEncoder import MyOrdinalEncoder
+
+from .preprocessing.my_scaler import standardscaling
+
+from .Column_Transformer import MyColumnTransformer
+from .my_pipeline import MyPipeline
+

varsha_preprocess-0.1.1/My_Preprocess1/encoders/MyLabelEncoder.py

@@ -0,0 +1,22 @@
+import numpy as np
+import pandas as pd
+
+class MyLabelEncoder:
+
+    def fit(self, y):
+        self.categories = np.unique(y)
+
+    def transform(self, y):
+        mapping_value ={val: index for index,val in enumerate(self.categories)} 
+        return y.map(mapping_value)
+    
+    def fit_transform(self, y):
+        self.fit(y)
+        return self.transform(y)
+            
+
+# df = pd.read_csv("Practice/covid_data.csv")
+# y = df['has_covid']
+
+# le = MyLabelEncoder()
+# print(le.fit_transform(y))

varsha_preprocess-0.1.1/My_Preprocess1/encoders/MyOneHotEncoder.py

@@ -0,0 +1,38 @@
+import numpy as np
+import pandas as pd
+
+class MyOneHotEncoder:
+
+    def fit(self, X):
+        self.categories = np.unique(X.dropna())
+
+    def transform(self, X):
+    
+        result = {}
+    
+        for category in self.categories:
+            # for this category, compare every value in X
+            # convert True-False to 1-0
+            # store in result dictionary with category as key
+            result[category] = (X == category).astype(int)
+        
+        # convert dictionary to DataFrame and return
+        return pd.DataFrame(result)
+
+    def fit_transform(self, X):
+        self.fit(X)
+        return self.transform(X)
+
+# df = pd.read_csv("Practice/covid_data.csv")
+# X = np.array(df['gender']).reshape(-1,1)
+
+# X = [['male']], [['female']]
+# df = pd.read_csv("Practice/SalaryData.csv")
+# X = df['Education Level']
+
+
+# print(type(X))
+# ohe = MyOneHotEncoder()
+# print(ohe.fit_transform(X))
+
+

varsha_preprocess-0.1.1/My_Preprocess1/encoders/MyOrdinalEncoder.py

@@ -0,0 +1,53 @@
+import numpy as np 
+import pandas as pd
+
+class MyOrdinalEncoder:
+    
+    # def ordinal_encoding(self, X):
+        
+    #     order = {
+    #         "Strong": 1,
+    #         "Mild": 0
+    #     }
+    #     #   we can use for loop but this is faster to do
+    #     return X.map(order)
+    
+    def __init__(self):
+        self.mapping = None
+        
+    def fit(self,X, order):
+        self.mapping = {}
+        
+        for i, category in enumerate(order):
+            self.mapping[category] = i
+            
+        self.mapping = self.mapping
+        return self
+
+    
+    def transform(self, X):
+        if self.mapping is None:
+            raise ValueError("first insert the order")
+        
+        result = []
+        for value in X:
+            if value in self.mapping is None:
+                result.append(self.mapping[value])
+                
+            else:
+                result.append(-1)
+     
+        return np.array(result).reshape(-1, 1)
+    
+    def fit_transform(self, X, order):
+        self.fit(X,order)
+        return self.transform(X)  
+    
+         
+# df = pd.read_csv("Practice/covid_data.csv")
+# X = df['cough']
+
+# oe = MyOrdinalEncoder()
+
+# print(oe.ordinal_encoding(X))
+

varsha_preprocess-0.1.1/My_Preprocess1/encoders/__init__.py

@@ -0,0 +1,3 @@
+from .MyLabelEncoder import MyLabelEncoder
+from .MyOneHotEncoder import MyOneHotEncoder
+from .MyOrdinalEncoder import MyOrdinalEncoder

varsha_preprocess-0.1.1/My_Preprocess1/my_pipeline.py

@@ -0,0 +1,23 @@
+from .encoders.MyLabelEncoder import MyLabelEncoder
+from .encoders.MyOneHotEncoder import MyOneHotEncoder
+from .encoders.MyOrdinalEncoder import MyOrdinalEncoder
+
+
+import numpy as np
+import pandas as pd
+
+class MyPipeline:
+    def __init__(self,steps):
+        self.steps = steps
+    
+    
+    def fit_transform(self,data):
+        for name, transformer in self.steps:
+            if len(self.steps)==0:
+                return data
+            data = transformer.fit_transform(data)
+            
+        return data
+    
+
+    

varsha_preprocess-0.1.1/My_Preprocess1/preprocessing/__init__.py

@@ -0,0 +1 @@
+from .my_scaler import standardscaling

varsha_preprocess-0.1.1/My_Preprocess1/preprocessing/my_scaler.py

@@ -0,0 +1,38 @@
+import numpy as np
+import pandas as pd
+
+class standardscaling:
+    def __init__(self):
+        self.mean = 0
+        self.standard_deviation = 0    
+        
+    def fit(self, X):
+        self.X = np.array(X).reshape(-1,1)
+        self.mean = np.mean(X, axis=0)
+        self.standard_deviation = np.std(X, axis =0)
+        return self.mean, self.standard_deviation
+    
+    def transform(self, X):
+        self.X = np.array(X).reshape(-1,1)
+        if np.all( self.standard_deviation == 0):
+            return (X-self.mean)
+        else:
+            return (X-self.mean)/self.standard_deviation
+        
+    
+    def fit_transform(self, X):
+        self.fit(X)
+        return self.transform(X)
+    
+
+# df = pd.read_csv("/home/varsha/PracticePython/100DaysOfMl/Practice/SalaryData.csv")
+
+# X = df['Age']
+# # print(X[3])
+# # X = [100,120,130,150]
+
+# standardscaler = standardscaling()
+# X_scaled = standardscaler.fit_transform(X)
+
+# print(X_scaled)
+

varsha_preprocess-0.1.1/PKG-INFO

@@ -0,0 +1,49 @@
+Metadata-Version: 2.4
+Name: varsha-preprocess
+Version: 0.1.1
+Summary: Custom preprocessing pipeline library
+Description-Content-Type: text/markdown
+Requires-Dist: numpy
+Requires-Dist: pandas
+Dynamic: description
+Dynamic: description-content-type
+Dynamic: requires-dist
+Dynamic: summary
+
+# My_Preprocess
+
+
+A lightweight machine learning preprocessing library built from scratch using NumPy and Pandas.
+
+This project is designed to deeply understand how core preprocessing tools like Pipeline, ColumnTransformer, and Encoders work internally — without relying on sklearn.
+
+---
+
+## ✨ Features
+
+- 🔁 Custom Pipeline (sequential transformations)
+- 🧩 ColumnTransformer (parallel column-wise transformations)
+- 🔤 Encoders:
+  - Label Encoder
+  - One Hot Encoder
+  - Ordinal Encoder
+- 📏 Standard Scaler
+- ⚙️ Fully built using NumPy and Pandas
+
+---
+
+## 🧠 Why this project?
+
+Most ML libraries (like sklearn) hide internal implementation details.
+
+This project focuses on:
+- Understanding how transformations are chained
+- Learning how fit / transform logic works
+- Building preprocessing systems from scratch
+
+---
+
+## 📦 Installation
+
+```bash
+pip install varsha-preprocess

varsha_preprocess-0.1.1/README.md

@@ -0,0 +1,37 @@
+# My_Preprocess
+
+
+A lightweight machine learning preprocessing library built from scratch using NumPy and Pandas.
+
+This project is designed to deeply understand how core preprocessing tools like Pipeline, ColumnTransformer, and Encoders work internally — without relying on sklearn.
+
+---
+
+## ✨ Features
+
+- 🔁 Custom Pipeline (sequential transformations)
+- 🧩 ColumnTransformer (parallel column-wise transformations)
+- 🔤 Encoders:
+  - Label Encoder
+  - One Hot Encoder
+  - Ordinal Encoder
+- 📏 Standard Scaler
+- ⚙️ Fully built using NumPy and Pandas
+
+---
+
+## 🧠 Why this project?
+
+Most ML libraries (like sklearn) hide internal implementation details.
+
+This project focuses on:
+- Understanding how transformations are chained
+- Learning how fit / transform logic works
+- Building preprocessing systems from scratch
+
+---
+
+## 📦 Installation
+
+```bash
+pip install varsha-preprocess

varsha_preprocess-0.1.1/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+

varsha_preprocess-0.1.1/setup.py

@@ -0,0 +1,19 @@
+from setuptools import setup, find_packages
+# from .My_Preprocess1.my_pipeline import MyPipeline
+
+
+with open("README.md", "r", encoding="utf-8") as f:
+    long_description = f.read()
+
+setup(
+    name="varsha-preprocess",
+    version="0.1.1",   
+    packages=find_packages(),
+    install_requires=["numpy", "pandas"],
+
+    description="Custom preprocessing pipeline library",
+
+    #  key part for readme
+    long_description=long_description,
+    long_description_content_type="text/markdown",
+)

varsha_preprocess-0.1.1/varsha_preprocess.egg-info/PKG-INFO

@@ -0,0 +1,49 @@
+Metadata-Version: 2.4
+Name: varsha-preprocess
+Version: 0.1.1
+Summary: Custom preprocessing pipeline library
+Description-Content-Type: text/markdown
+Requires-Dist: numpy
+Requires-Dist: pandas
+Dynamic: description
+Dynamic: description-content-type
+Dynamic: requires-dist
+Dynamic: summary
+
+# My_Preprocess
+
+
+A lightweight machine learning preprocessing library built from scratch using NumPy and Pandas.
+
+This project is designed to deeply understand how core preprocessing tools like Pipeline, ColumnTransformer, and Encoders work internally — without relying on sklearn.
+
+---
+
+## ✨ Features
+
+- 🔁 Custom Pipeline (sequential transformations)
+- 🧩 ColumnTransformer (parallel column-wise transformations)
+- 🔤 Encoders:
+  - Label Encoder
+  - One Hot Encoder
+  - Ordinal Encoder
+- 📏 Standard Scaler
+- ⚙️ Fully built using NumPy and Pandas
+
+---
+
+## 🧠 Why this project?
+
+Most ML libraries (like sklearn) hide internal implementation details.
+
+This project focuses on:
+- Understanding how transformations are chained
+- Learning how fit / transform logic works
+- Building preprocessing systems from scratch
+
+---
+
+## 📦 Installation
+
+```bash
+pip install varsha-preprocess

varsha_preprocess-0.1.1/varsha_preprocess.egg-info/SOURCES.txt

@@ -0,0 +1,16 @@
+README.md
+setup.py
+My_Preprocess1/Column_Transformer.py
+My_Preprocess1/__init__.py
+My_Preprocess1/my_pipeline.py
+My_Preprocess1/encoders/MyLabelEncoder.py
+My_Preprocess1/encoders/MyOneHotEncoder.py
+My_Preprocess1/encoders/MyOrdinalEncoder.py
+My_Preprocess1/encoders/__init__.py
+My_Preprocess1/preprocessing/__init__.py
+My_Preprocess1/preprocessing/my_scaler.py
+varsha_preprocess.egg-info/PKG-INFO
+varsha_preprocess.egg-info/SOURCES.txt
+varsha_preprocess.egg-info/dependency_links.txt
+varsha_preprocess.egg-info/requires.txt
+varsha_preprocess.egg-info/top_level.txt

varsha_preprocess-0.1.1/varsha_preprocess.egg-info/dependency_links.txt

@@ -0,0 +1 @@
+

varsha_preprocess-0.1.1/varsha_preprocess.egg-info/requires.txt

@@ -0,0 +1,2 @@
+numpy
+pandas

varsha_preprocess-0.1.1/varsha_preprocess.egg-info/top_level.txt

@@ -0,0 +1 @@
+My_Preprocess1