Open In App

Implementation of Radius Neighbors from Scratch in Python

Improve
Improve
Like Article
Like
Save
Share
Report

Radius Neighbors is also one of the techniques based on instance-based learning. Models based on instance-based learning generalize beyond the training examples. To do so, they store the training examples first. When it encounters a new instance (or test example), then they instantly build a relationship between stored training examples and this new instant to assign a target function value for this new instance. Instance-based methods are sometimes called lazy learning methods because they postponed learning until the new instance is encountered for prediction.

Instead of estimating the hypothetical function (or target function) once for the entire space,  these methods will estimate it locally and differently for each new instance to be predicted.

Radius Neighbors Classifier:

Basic Assumptions

  1. All instances correspond to points in the n-dimensional space where n represents the number of features in any instance.
  2. The neighbors of an instance are defined in terms of the Euclidean distance.
An instance can be represented by < x1, x2, .............., xn >.
Euclidean distance between two instances xa and xb is given by d( xa, xb ) : 

\sqrt{\sum_{j=1}^{n}\left(x_{j}^{a}-x_{j}^{b}\right)^{2}}
 

Euclidean Distance 


How does it work?


 

Radius Neighbors Classifier first stores the training examples. During prediction, when it encounters a new instance ( or test example) to predict, it finds the number of neighbors from training instances within a fixed radius of r center at test instance, where r is a floating-point value specified by the user.  Then assigns the most common class among the training instances within that radius to the test instance.


 

The optimal choice for r is by validating errors on test data.


 

Radius Neighbors Classification Graphical Representation


 

In the above figure, “+” denotes training instances labelled with 1. “-” denotes training instances with 0. Here we classified for the test instance xt as the most common class among training instances within the circle. Here, r is a user-specified choice. In the above figure, positive instances are in majority in the circle, so xt is classified as “+” or 1.


 

Pseudocode:

  1. Store all training instances.
  2. Repeat steps 3, 4, and 5 for each test instance.
  3. Finds the number of neighbors from training instances within a fixed radius of r center at a test instance.
  4. y_pred for current test example =  most common class among training instances within a circle.
  5. Go to step 2.

Implementation


 


 

It has 8 features columns like i.e “Age”, “Glucose” e.t.c, and the target variable “Outcome” for 108 patients. So in this, we will create a K-Nearest Neighbors Classifier model to predict the presence of diabetes or not for patients with such information.


 

Python3

# Importing libraries
  
import pandas as pd
  
import numpy as np
  
from sklearn.model_selection import train_test_split
  
from scipy.stats import mode
  
from sklearn.neighbors import RadiusNeighborsClassifier
  
# Radius Nearest Neighbors Classification
  
class Radius_Nearest_Neighbors_Classifier() : 
      
    def __init__( self, r ) :
          
        self.r = r
          
    # Function to store training set
          
    def fit( self, X_train, Y_train ) :
          
        self.X_train = X_train
          
        self.Y_train = Y_train
          
        # no_of_training_examples, no_of_features
          
        self.m, self.n = X_train.shape
      
    # Function for prediction
          
    def predict( self, X_test ) :
          
        self.X_test = X_test
          
        # no_of_test_examples, no_of_features
          
        self.m_test, self.n = X_test.shape
          
        # initialize Y_predict
          
        Y_predict = np.zeros( self.m_test )
          
        for i in range( self.m_test ) :
              
            x = self.X_test[i]
              
            # find the number of neighbors within a fixed 
            # radius r of current training example
              
            neighbors = self.find_neighbors( x )
              
            # most frequent class in the circle drawn by current 
            # training example of fixed radius r
              
            Y_predict[i] = mode( neighbors )[0][0]
              
        return Y_predict
      
    # Function to find the number of neighbors within a fixed radius 
    # r of current training example
            
    def find_neighbors( self, x ) :
          
        # list to store training examples which will fall in the circle
          
        inside = []
          
        for i in range( self.m ) :
              
            d = self.euclidean( x, self.X_train[i] )
              
            if d <= self.r :
                  
                inside.append( self.Y_train[i] )
  
        inside_array = np.array( inside )
                  
        return inside_array
      
    # Function to calculate euclidean distance
              
    def euclidean( self, x, x_train ) :
          
        return np.sqrt( np.sum( np.square( x - x_train ) ) )
  
 # driver code
  
def main() :
      
    # Create dataset
      
    df = pd.read_csv( "diabetes.csv" )
  
    X = df.iloc[:,:-1].values
  
    Y = df.iloc[:,-1:].values
      
    # Splitting dataset into train and test set
  
    X_train, X_test, Y_train, Y_test = train_test_split( 
      X, Y, test_size = 1/3, random_state = 0 )
      
    # Model training
      
    model = Radius_Nearest_Neighbors_Classifier( r = 550 )
      
    model.fit( X_train, Y_train )
      
    model1 = RadiusNeighborsClassifier( radius = 550 )
      
    model1.fit( X_train, Y_train )
      
    # Prediction on test set
  
    Y_pred = model.predict( X_test )
      
    Y_pred1 = model1.predict( X_test )
      
    # measure performance
      
    correctly_classified = 0
      
    correctly_classified1 = 0
      
    # counter
      
    count = 0
      
    for count in range( np.size( Y_pred ) ) :
          
        if Y_test[count] == Y_pred[count] :
              
            correctly_classified = correctly_classified + 1
              
        if Y_test[count] == Y_pred1[count] :
              
            correctly_classified1 = correctly_classified1 + 1
          
        count = count + 1
          
    print("Accuracy on test set by our model     : ", ( 
      correctly_classified / count ) * 100 )
      
    print("Accuracy on test set by sklearn model : ", ( 
      correctly_classified / count ) * 100 )
  
   
if __name__ == "__main__"
      
    main()

                    

 
 

Output  :

Accuracy on test set by our model     :  61.111111111111114
Accuracy on test set by sklearn model :  61.111111111111114


 

The accuracy achieved by our model and sklearn is equal which indicates the correct implementation of our model.


 

Note: Above Implementation is for model creation from scratch, not to improve the accuracy of the diabetes dataset.


 

Radius Neighbors Regressor:

Radius Neighbors Regressor first stores the training examples. During prediction, when it encounters a new instance ( or test example ) to predict,  it finds the number of neighbors from training instances within a fixed radius of r center at test instance, where r is a floating-point value specified by the user.  Then assigns the mean of the training instances within that radius to the test instance.


 

The optimal choice for r is by validating errors on test data.


 

Pseudocode:

  1. Store all training instances.
  2. Repeat steps 3, 4, and 5 for each test instance.
  3. Finds the number of neighbors from training instances within a fixed radius of r center at a test instance.
  4. y_pred for current test example =  mean of the training instances within a circle.
  5. Go to step 2.

Implementation:


 

Dataset used in this implementation can be downloaded from link.


 

It has 2 columns — “YearsExperience” and “Salary” for 30 employees in a company. So in this, we will create a Radius Neighbors Regression model to learn the correlation between the number of years of experience of each employee and their respective salary.


 

The model, we created predicts the same value as the sklearn model predicts for the test set.


Code: 
 

Python3

# Importing libraries
  
import pandas as pd
import numpy as np
from sklearn.model_selection import train_test_split
from sklearn.neighbors import RadiusNeighborsRegressor
  
# Radius Nearest Neighbors Regression
  
class Radius_Nearest_Neighbors_Regression() : 
      
    def __init__( self, r ) :
          
        self.r = r
          
    # Function to store training set
          
    def fit( self, X_train, Y_train ) :
          
        self.X_train = X_train
          
        self.Y_train = Y_train
          
        # no_of_training_examples, no_of_features
          
        self.m, self.n = X_train.shape
      
    # Function for prediction
          
    def predict( self, X_test ) :
          
        self.X_test = X_test
          
        # no_of_test_examples, no_of_features
          
        self.m_test, self.n = X_test.shape
          
        # initialize Y_predict
          
        Y_predict = np.zeros( self.m_test )
          
        for i in range( self.m_test ) :
              
            x = self.X_test[i]
              
            # find the number of neighbors within a fixed 
            # radius r of current training example
              
            neighbors = self.find_neighbors( x )
              
            # mean of the neighbors in the circle drawn by 
            # current training example of fixed radius r
              
            Y_predict[i] = np.mean( neighbors )
              
        return Y_predict
      
    # Function to find the number of neighbors within a fixed 
    # radius r of current training example
            
    def find_neighbors( self, x ) :
          
        # list to store training examples which will fall in the circle
          
        inside = []
          
        for i in range( self.m ) :
              
            d = self.euclidean( x, self.X_train[i] )
              
            if d <= self.r :
                  
                inside.append( self.Y_train[i] )
  
        inside_array = np.array( inside )
                  
        return inside_array
      
    # Function to calculate euclidean distance
              
    def euclidean( self, x, x_train ) :
          
        return np.sqrt( np.sum( np.square( x - x_train ) ) )
        
  
# driver code
  
def main() :
      
    # Importing dataset
      
    df = pd.read_csv( "salary_data.csv" )
  
    X = df.iloc[:,:-1].values
  
    Y = df.iloc[:,1].values
      
    # Splitting dataset into train and test set
  
    X_train, X_test, Y_train, Y_test = train_test_split( 
      X, Y, test_size = 1/3, random_state = 0 )
      
    # Model training
      
    model = Radius_Nearest_Neighbors_Regression( r = 550 )
      
    model.fit( X_train, Y_train )
      
    model1 = RadiusNeighborsRegressor( radius = 550 )
  
    model1.fit( X_train, Y_train )
      
    # Prediction on test set
      
    Y_pred = model.predict( X_test )
  
    Y_pred1 = model1.predict( X_test )
      
    print( "Real values                         : ", Y_test[:3] )
      
    print( "Predicted values by our model       : ", np.round( Y_pred[:3], 2 ) )
      
    print( "Predicted values by sklearn model   : ", np.round( Y_pred1[:3], 2 ) )
     
  
if __name__ == "__main__"
      
    main()

                    

Output:

Real values                         :  [ 37731 122391  57081]
Predicted values by our model       :  [71022.5 71022.5 71022.5]
Predicted values by sklearn model   :  [71022.5 71022.5 71022.5]


Last Updated : 21 Mar, 2024
Like Article
Save Article
Previous
Next
Share your thoughts in the comments
Similar Reads