lacli

Linear algebra in your CLI — yayo!

LAcli runs in your terminal and provides a set of linear algebra commands. All operations are performed using NumPy.

The Problem

How can be possible to support an huge number of filetype, format and standard without fill the software with useless dependencies for the final user, given that in the daily work use only a bunch of filetype ?

Hard, semi-stupid but cool solution

To support a very big number of filetype, instead to implement a complete parser or worse a huge library for any filetype, read the raw file, every time you encounter a number, consider that number a part of the matrix or the math object, and after based on the file type, delete the unusefull number, or something like that

More easy and pratical solution

To support a veeery big number of filetype, without fill the software with a huge number of dependencies for parsing files, the program can download the parser in-time at runtime, the fact is that the final user need the be able to work with just a bunch a filetype, so the most of eventual dependencies are useless, so, instead the principle to use only what you need is a good thing, the downside is the time spent in download, but in this case a local cache can be usefull to delimiting the download only to the first time

---

Technology Stack Problem

Pure Python is too slow, approximatelly 3sec for cicle 10000 numbers, inaceptable counting the opening of the file in memory

Change language for file loading and matrix creation

Go ? C ? C++ ? boh (boh for now)

I/O

You can set stdin and stdout to use files, pipes, or both in powerful combinations.

Argument Parser

The argument parser should be devided by commands, like BMO (basic matrix operations), checks, rotation, factorization and so on, Each command should be devided by single feature, so every final feature should be accessible via a command and not by direct access, Nothing forbids to organize the command accesible feature in sub-commands and so on

Supported File Types

General Purpose

Format	Extensions	Supported
JSON	.json	NO
CSV	.csv	NO

Multidimensional & Big Data Formats

Format	Extensions	Description	Supported
Numpy	.npy,.npz	A simple format for saving numpy arrays to disk with the full information about them	NO
Matlab	.mat	The official matlab file format to store array and general purpose data	NO
HDF5	.h5, .hdf5	Hierarchical Data Format. Essential for storing massive, highly complex, and multidimensional datasets	NO
NetCDF	.nc	Network Common Data Form. Heavily utilized in atmospheric, oceanic, and meteorological sciences to store array-oriented scientific data	NO
Matrix Market	.mtx	The Matrix Market (MM) exchange formats provide a simple mechanism to facilitate the exchange of matrix data	NO
PyTorch	.pt, .pth	The official PyTorch format to store model weights and architecture	NO
GGML	.ggml	GGML is an open-source tensor library written in C and C++ designed for running machine learning models	NO
GGUF	.gguf	GGUF (GPT-Generated Unified Format) is a highly popular file format designed to store and run large language models (LLMs)	NO
ONNX	.onnx	ONNX is an open format built to represent machine learning model	NO

Bioinformatics & Genomics

Format	Extensions	Description	Supported
FASTA/FASTQ	.fasta, .fastq	Industry standards for genomic data	NO
BAM/SAM	.bam, .sam	Compressed binary formats used to store sequence alignments mapped to a reference genome	NO

Chemistry & Material Science

Format	Extensions	Description	Supported
PDB	.pdb	Protein Data Bank. Used to store 3D structural data for biological macromolecules, such as proteins and nucleic acids	NO
SDF	.sdf	Structure Data File. Contains chemical structures and associated tabular or numerical properties	NO

Tabular Data & Statistics

Format	Extensions	Description	Supported
Parquet	.parquet	Highly efficient, columnar storage used extensively in data science	NO

---

TODO:

[1] The first big step for the development of this project is to add the support for a cospicious number of format file, so the user need to be able to compute some basic operation

• [1.1] Implement the argument parser with the BSO command and some feature to test the file implementation
• [1.2] Implement support for JSON
• [1.3] Implement support for CSV

Feature

[0] Basic Matrix Operations

• [0.1] matrix mul (arbitrary dimensions)
• [0.2] matrix sum
• [0.3] matrix scalar mul
• [0.4] scalar dot product
• [0.5] scalar sum
• [0.6] inverse compute
• [0.7] transpose compute
• [0.8] rank compute

[1] Checks

• [1.1] invertability check
• [1.2] vectors independence check
• [1.3] vectors orthogonality check
• [1.4] symmetric check
• [1.5] triangular check
• [1.6] positive definite check

[2] Rotations

• [2.1] rotation matrix computation (input: matrix, angle, [axis], [center], [scale], [shear], [perspective])

[3] Factorization

• [3.1] gauss-jordan elimination
• [3.2] LU decomposition
• [3.3] LDU decomposition
• [3.4] QR decomposition
• [3.5] Cholesky decomposition
• [3.6] orthogonal decomposition
• [3.7] SVD
• [3.8] eigenvalue decomposition

[4] Least Squares

• [4.1] least squares
• [4.2] weighted least squares
• [4.3] least squares with regularization
• [4.4] regularization
• [4.5] linear regression

[5] I/O Support

The support word below means support for input and output (read/write) in combinations between file format

• [5.1] support for JSON
• [5.2] support for CSV
• [5.3] support for CLI
• [5.4] support for Parquet
• [5.5] support for NetCDF
• [5.6] support for FASTA/FASTQ
• [5.7] support for HDF5
• [5.8] support for SDF
• [5.9] support for PDB
• [5.10] support for Numpy
• [5.11] support for Matlab
• [5.12] support for GGML
• [5.13] support for GGUF
• [5.14] support for ONNX
• [5.15] support for Matrix Market
• [5.16] support for PyTorch