A Python library for data transformations
Project description
Data Transformations Library
This is a Python library for performing common data transformations used in machine learning and data science workflows. The library includes functions for transposing matrices, creating time series windows, and performing 2D cross-correlation.
Features
- Transpose Function: Transposes a 2D matrix (list of lists).
- Time Series Windowing Function: Creates windows from a 1D array with specified size, shift, and stride.
- Cross-Correlation Function: Performs 2D cross-correlation on an input matrix using a given kernel.
Installation
To install the library, use pip3
:
pip3 install eimantas_data_transformations
Or, if you are using Poetry, add it to your project with:
poetry add eimantas_data_transformations
Usage
Transpose Function
Transposes a 2D matrix (list of lists). This function takes a 2D matrix represented as a list of lists and returns a new matrix that is the transpose of the input matrix. Transposing a matrix means switching the rows and columns.
Example:
from eimantas_data_transformations.transformations import transpose2d
matrix = [[1, 2, 3], [4, 5, 6]]
transposed = transpose2d(matrix)
print(transposed)
# Output: [[1, 4], [2, 5], [3, 6]]
Time Series Windowing Function
Creates windows from a 1D array with specified size, shift, and stride.
Example:
from eimantas_data_transformations.transformations import window1d
input_array = [1, 2, 3, 4, 5, 6, 7, 8, 9]
windows = window1d(input_array, size=3, shift=2, stride=2)
print(windows)
# Output: [[1, 3, 5], [3, 5, 7], [5, 7, 9]]
Cross-Correlation Function
Performs 2D cross-correlation on an input matrix using a given kernel.
Example:
import numpy as np
from eimantas_data_transformations.transformations import convolution2d
input_matrix = np.array([
[1, 2, 3, 4],
[5, 6, 7, 8],
[9, 10, 11, 12],
[13, 14, 15, 16]
])
kernel = np.array([
[1, 0],
[0, 1]
])
result = convolution2d(input_matrix, kernel, stride=2)
print(result)
# Output: [[ 7. 11.]
# [23. 27.]]
Author
Eimantas Venslovas
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for eimantas_data_transformations-0.1.1.tar.gz
Algorithm | Hash digest | |
---|---|---|
SHA256 | f2b3e687000fce318fb0f0fdb27b974737641605b8dd962c48481b9545bb5027 |
|
MD5 | 2590d224249231db7d48bfa5c5cd4c9f |
|
BLAKE2b-256 | 76571d6732162d4e04a1df4d0f19737e130ec46e3c36b600255683df8e950d7c |
Hashes for eimantas_data_transformations-0.1.1-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 1a81f2c4aa7a3eea24e82cc659f7a60ddee0ed85dbac497b04ede153e295f978 |
|
MD5 | 38aff52c787f0d4d444e3d7718c344a1 |
|
BLAKE2b-256 | b65dc459c3da711cee8118190f2271acb267ecca9ad527b8a93c3bfe14f8c5ad |