Merge pull request #4 from fszewczyk/doxygen

Micrograd implementation

.gitignore

@@ -1,6 +1,9 @@
 build/
 data/
 
+docs/html
+docs/latex
+
 .vscode/
 .clang-format
 *.out

README.md

@@ -1,110 +1,3 @@
 <div align="center">
 
 <h1>Shkyera Tensor</h1>
-
-<div>
-
-![LinuxBuild](https://github.com/fszewczyk/shkyera-tensor/actions/workflows/linux.yml/badge.svg) 
-![MacOSBuild](https://github.com/fszewczyk/shkyera-tensor/actions/workflows/macos.yml/badge.svg) 
-![MacOSBuild](https://github.com/fszewczyk/shkyera-tensor/actions/workflows/windows.yml/badge.svg) 
-[![LICENSE](https://img.shields.io/badge/license-Beerware-yellow)](LICENSE) 
-
-</div>
-<i>First Deep Learning library <b>without</b> an implementation of a tensor</i>
-</div>
-
-## Usage
-See `/examples` for more detailed usage.
-
-### Quick start
-Simply import the library in your code
-```cpp
-#include "include/ShkyeraTensor.hpp"
-```
-
-### Dataset
-```cpp
-// The first column is assumed to be the labels
-// Labels are catgorical
-st::Dataset<int, double> data("tests/preview.csv", CATEGORICAL); 
-
-// Shuffles the data
-data.shuffle();            
-
-// Divides dataset into train and test set, with 0.6 proportion                         
-std::pair<st::Dataset<int,double>, st::Dataset<int,double>> trainAndTest = data.splitIntoTrainAndTest(0.6); 
-
-// Gets first 5 rows of the Dataset
-st::Dataset<int,double> headTestData = trainAndTest.second.head(); 
-```
-
-### Network
-```cpp
-// Creating a neural network with 3 hidden layers
-// First one has 50 neurons and uses sigmoid activation
-// Second one has 30 neurons and uses sigmoid activation
-// First one has 10 neurons and uses ReLU activation (default one)
-st::Network nn(data, {50, 30, 10}, {st::SIGMOID, st::TANH});
-
-// Train the network with 0.01 learning rate, batch size of 64 for 10 epochs
-nn.train(0.01, 64, 10);
-```
-
-### Linear Algebra
-#### Vector
-```cpp
-// Vector of 3 integers
-st::Vector<int> vInt(3); 
-
-// Vector of size 3, filled with 2.5 doubles
-st::Vector<double> vDouble(3,2.5); 
-
-// Float Vector of size 5, filled with random values between -3 and 2
-auto randomFloatVec = st::Vector<float>::random(5, -3, 2); 
-
-// Zero Vector of size 4
-auto zerosIntVec = st::Vector<int>::zeros(4); 
-
-// Sample algebra operations
-st::Vector<float> resultDot = vDouble.dot(vInt);
-resultDot *= 2; 
-resultDot -= 5; 
-```
-
-#### Matrix
-```cpp
-// 4x3 integer matrix filled with 2
-st::Matrix<int> twosMat(4, 3, 2);          
-
-// 5x4 matrix with random doubles between -2 and 1
-auto randomMat = st::Matrix<double>::random(5, 4, -2, -1); 
-
-// 5x4 matrix with random ones
-auto onesMat = st::Matrix<double>::ones(4, 2);             
-
-// Sample algebra operations
-auto dotProduct = randomMat.dot(st ::Vector<double>(5, 1)); 
-auto matrixMultiplication = randomMat * onesMat;           
-randomMat *= onesMat; 
-randomMat += 3;
-randomMat *= 2;
-```
-
-## Roadmap
-- [X] Basic Linear Algebra
-- [X] Loading `.csv` datasets
-- [ ] Basic Machine Learning
-- [ ] Basic MLP
-- [ ] Tensor
-- [ ] Simple Reccurent Layers
-- [ ] Echo State Networks
-- [ ] LSTM Layer
-- [ ] Transformers
-- [ ] Convolutional Neural Networks
-
-## Contributing
-### Formatting
-Make sure you use the appropriate formatting settings. Simple run this in the main directory:
-```
-clang-format -style='{ColumnLimit: 120, IndentWidth: 4}' -dump-config > .clang-format
-```