mikoshilang 0.3.0

Symbolic computation language for Python — pattern matching, algebraic simplification, and computational intelligence

MikoshiLang

A symbolic computation language for Python — Wolfram-style syntax, pattern matching, algebraic simplification, and domain-specific packages for physics, chemistry, and signal processing.

Built by Mikoshi Ltd.

Why MikoshiLang?

Feature	MikoshiLang	SymPy	Wolfram
Wolfram-style syntax (Sin[x], {1,2,3})	✅	❌	✅
Pattern matching (x_, __, conditions)	✅	Limited	✅
Rule-based rewriting engine	✅	Limited	✅
Interactive REPL with In/Out history	✅	❌	✅
Jupyter kernel with LaTeX rendering	✅	✅	✅
Chemistry — 118 elements, equation balancing	✅	❌	✅
Physics units with arithmetic & conversion	✅	✅	✅
Signal processing (FFT, filters, spectrograms)	✅	Limited	✅
Free & open source	✅	✅	❌ ($395/yr)
Python-native, pip installable	✅	✅	❌

Key Selling Points

• 🧪 Chemistry built-in — All 118 elements with atomic mass, electron configuration, electronegativity. Balance equations: BalanceEquation["H2 + O2 -> H2O"] → "2H2 + O2 -> 2H2O". Calculate molecular mass: MolecularMass["C6H12O6"] → 180.156
• ⚡ Wolfram syntax, Python ecosystem — Write Solve[x^2 - 4 == 0, x] not sympy.solve(sympy.Symbol('x')**2 - 4, sympy.Symbol('x')). Same power, 70% less typing
• 🎯 Pattern matching — Real Wolfram-style patterns: f[x_] := x^2, blanks, sequences, conditions. Not regex — structural matching on expression trees
• 📡 Signal processing — DFT, filters (low/high/band-pass), convolution, window functions, spectrograms — all from one import
• 🔬 Physics units — 50+ units, quantity arithmetic that checks dimensions, automatic conversion: UnitConvert[Quantity[100, "cm"], "m"]
• 📓 Jupyter kernel — LaTeX-rendered expressions, inline plots, proper notebook experience

Installation

pip install mikoshilang

# With Jupyter support
pip install mikoshilang[jupyter]

# With signal processing
pip install mikoshilang[signal]

# Everything
pip install mikoshilang[all]

Language Syntax

MikoshiLang uses Wolfram-style syntax. Launch the REPL:

mikoshilang

Arithmetic

In[1]:= 2 + 3 * x
Out[1]= 2 + 3*x

In[2]:= x^2 - 4
Out[2]= x^2 - 4

In[3]:= (x + 1)(x - 1)    (* implicit multiplication *)
Out[3]= (x + 1)*(x - 1)

Function Calls (Square Brackets)

In[1]:= Sin[Pi/2]
In[2]:= Diff[x^2, x]
In[3]:= Integrate[x^2, x]
In[4]:= Solve[x^2 - 4 == 0, x]
In[5]:= Simplify[(x^2 - 1)/(x - 1)]
In[6]:= Factor[x^2 - 4]
In[7]:= Expand[(x + 1)^3]
In[8]:= Limit[Sin[x]/x, x -> 0]
In[9]:= Series[Exp[x], {x, 0, 5}]

Lists and Data

In[1]:= {1, 2, 3, 4, 5}
In[2]:= Range[10]
In[3]:= Table[i^2, {i, 1, 10}]
In[4]:= Map[Sin, {1, 2, 3}]
In[5]:= Select[{1, -2, 3, -4}, Positive]

Matrices

In[1]:= Det[{{1, 2}, {3, 4}}]
Out[1]= -2

In[2]:= Inverse[{{1, 2}, {3, 4}}]

Pattern Matching and Rules

In[1]:= x /. x -> 3
In[2]:= f[x_] := x^2
In[3]:= MatchQ[Sin[x], Sin[_]]
In[4]:= ReplaceAll[x + y, {x -> 1, y -> 2}]

Constants

Pi, E, I, Infinity, True, False

Comments

(* This is a comment *)

Jupyter Integration

Install the kernel:

pip install mikoshilang[jupyter]
python -m mikoshilang.jupyter.install

Then open Jupyter Notebook and select the "MikoshiLang" kernel. Features:

• LaTeX rendering of expressions
• Inline matplotlib plots with Plot[Sin[x], {x, -Pi, Pi}]
• Rich display of matrices and lists

Plotting

Plot[Sin[x], {x, -Pi, Pi}]
Plot[{Sin[x], Cos[x]}, {x, -Pi, Pi}]
ListPlot[{1, 4, 9, 16, 25}]
ListLinePlot[{1, 4, 9, 16, 25}]

Physics Units

In[1]:= q = Quantity[9.8, "m/s^2"]
In[2]:= t = Quantity[3, "s"]
In[3]:= q * t
Out[3]= Quantity[29.4, "m/s"]

In[4]:= UnitConvert[Quantity[100, "cm"], "m"]
Out[4]= Quantity[1, "m"]

In[5]:= UnitConvert[Quantity[72, "kg"], "lb"]
Out[5]= Quantity[158.73, "lb"]

Supported Units

Category	Units
Length	m, cm, mm, km, in, ft, yd, mi
Mass	kg, g, mg, lb, oz
Time	s, ms, min, h, day
Speed	m/s, km/h, mph
Force	N, lbf
Energy	J, kJ, cal, kcal, eV, kWh
Power	W, kW, hp
Pressure	Pa, kPa, atm, bar, psi
Temperature	K, C, F
Electric	A, V, ohm, Farad, H, Coulomb, Hz

Physical Constants

SpeedOfLight, GravitationalConstant, PlanckConstant, BoltzmannConstant, AvogadroNumber, ElementaryCharge

Chemistry

In[1]:= Element["H"]
Out[1]= {name: "Hydrogen", number: 1, mass: 1.008, symbol: "H"}

In[2]:= AtomicMass["O"]
Out[2]= 15.999

In[3]:= ElectronConfiguration["Fe"]
Out[3]= [Ar] 3d6 4s2

In[4]:= MolecularMass["H2O"]
Out[4]= 18.015

In[5]:= MolecularMass["C6H12O6"]
Out[5]= 180.156

In[6]:= BalanceEquation["H2 + O2 -> H2O"]
Out[6]= 2H2 + O2 -> 2H2O

All 118 elements included with atomic number, symbol, name, mass, electron configuration, electronegativity, and category.

Signal Processing

Requires pip install mikoshilang[signal] for filters and spectrogram.

In[1]:= DFT[{1, 2, 3, 4}]
In[2]:= IDFT[{10, -2, -2, -2}]

In[3]:= Convolve[{1, 2, 3}, {0, 1, 0.5}]

In[4]:= HammingWindow[256]
In[5]:= HanningWindow[256]
In[6]:= BlackmanWindow[256]

(* Symbolic Fourier transforms via SymPy *)
In[7]:= FourierTransform[Exp[-t^2], t, w]

(* Filters (require scipy) *)
In[8]:= LowPassFilter[data, cutoff]
In[9]:= HighPassFilter[data, cutoff]
In[10]:= BandPassFilter[data, low, high]

In[11]:= Spectrogram[data, sample_rate]

Python API

from mikoshilang import *

# Parse and evaluate Wolfram-style syntax
result = parse_and_eval("Simplify[(x^2 - 1)/(x - 1)]")

# Or use Python constructors directly
x = Symbol("x")
expr = x**2 + 2*x + 1
print(simplify(expr))
print(to_latex(expr))

# Units
q = Quantity(100, "cm")
print(UnitConvert(q, "m"))

# Chemistry
print(MolecularMass("C6H12O6"))
print(BalanceEquation("H2 + O2 -> H2O"))

Feature Comparison with Wolfram

Feature	Wolfram	MikoshiLang
Symbolic algebra	✅	✅ (via SymPy)
Pattern matching	✅	✅
Calculus	✅	✅
Linear algebra	✅	✅ (via NumPy)
Number theory	✅	✅
Wolfram-style syntax	✅	✅
Jupyter notebooks	✅	✅
LaTeX output	✅	✅
Plotting	✅	✅ (via Matplotlib)
Physics units	✅	✅
Chemistry	✅	✅ (118 elements)
Signal processing	✅	✅ (via SciPy)
Free & open source	❌	✅

License

Apache 2.0 — Mikoshi Ltd.