Cumulative Distribution Transform Nearest Subspace (CDT-NS) Classifier

This tutorial will demonstrate how to use the CDT-NS classifier for 1D data in the PyTransKit package.

Class:: CDT_NS

Functions:

1. Constructor function: cdt_ns_obj = CDT_NS(num_classes, rm_edge)

   Inputs:
   ----------------
   num_classes : integer value
       totale number of classes in the dataset.
   rm_edge : boolean
       IF TRUE the first and last points of CDTs will be removed.
   
   Outputs:
   ----------------
   cdt_ns_obj : class object
       Instance of the class CDT_NS.
   

2. Fit function: cdt_ns_obj.fit(Xtrain, Ytrain, no_deform_model)

   Inputs:
   ----------------
   Xtrain : array-like, shape (n_samples, n_columns)
       1D data for training.
   Ytrain : ndarray of shape (n_samples,)
       Labels of the training samples.
   no_deform_model : boolean flag; IF TRUE, no deformation model will be added
       default = False.
   

3. Predict function: preds = cdt_ns_obj.predict(Xtest, use_gpu)

   Inputs:
   ----------------
   Xtest : array-like, shape (n_samples, n_columns)
       1D data for testing.
   use_gpu: boolean flag; IF TRUE, use gpu for calculations
       default = False.
   
   Outputs:
   ----------------
   preds : 1d array, shape (n_samples,)
      Predicted labels for test samples.
   

Example

The following example will demonstrate how to: * create and initialize an instance of the class CDT_NS * train the model with training 1D samples * apply the model to predict calss labels of the test 1D samples In this example we have used a synthetic dataset (1D) stored in the data folder. The dataset contains two classes. Class 0: different translated versions of Gaussian signal Class 1: translated versions of summation of two Gaussian signals

Import some python libraries

import numpy as np
from sklearn.metrics import accuracy_score
from pathlib import Path
import sys
sys.path.append('../')
from pytranskit.classification.utils import *

use_gpu = False

Import CDT-NS class from PyTransKit package

from pytranskit.classification.cdt_ns import CDT_NS

Load dataset

For loading data we have used load_data_1D function from the pytranskit/classifier/utils.py script. It takes name and directory of the dataset, and total number of classes as input. Returns both train and test samples in two separate 2d arrays of shape (n_samples, n_columns), and corresponding class labels. User can use there own implementation to load data, just need to make sure that the output arrays are consistent.

datadir = './data'
dataset = 'synthetic_1D'
num_classes = 2          # total number of classes in the dataset
(x_train, y_train), (x_test, y_test) = load_data_1D(dataset, num_classes, datadir)  # load_data function from utils.py

loading data from mat files
x_train.shape (1400, 201) x_test.shape (600, 201)
saved to ./data/synthetic_1D/dataset.hdf5

In this example we have used 512 randomly chosen samples per class to train the model. We have used another function take_train_samples function from utils.py script for this. User can use their own script.

n_samples_perclass = 512  # total number of training samples per class used in this example
x_train_sub, y_train_sub = take_train_samples(x_train, y_train, n_samples_perclass,
                                              num_classes, repeat=0) # function from utils.py

Create an instance of CDT_NS class

cdt_ns_obj = CDT_NS(num_classes, rm_edge=True)

Training phase

This function takes the train samples and labels as input, and stores the basis vectors for corresponding classes in a private variable. This variable will be used in the predict function in the test phase

print(x_train_sub.shape)
cdt_ns_obj.fit(x_train_sub, y_train_sub)

(1024, 201)

Calculating CDTs for training data ...
Generating basis vectors for each class ...

Testing phase

predict function takes the train samples as input and returns the predicted class labels

preds = cdt_ns_obj.predict(x_test, use_gpu)

Calculating CDTs for testing samples ...
Finding nearest subspace for each test sample ...

print('\nTest accuracy: {}%'.format(100*accuracy_score(y_test, preds)))

Test accuracy: 100.0%