How can I generate a matrix A with n rows and 2 columns in SymPy where n is a symbolic variable in Python?
import sympy
n = sympy.Symbol("n")
A = sympy.Matrix.zeros(n,2)
It went wrong.
TypeError: cannot determine truth value of Relational
I want to solve this problem and make it possible to generate a matrix like this.
To generate a matrix with a symbolic number of rows and/or columns you can use FunctionMatrix like this:
from sympy import symbols, FunctionMatrix, Lambda
i, j, m, n = symbols('i j m n', integer=True)
A = FunctionMatrix(m, n,Lambda((i, j), 0))
A[100,100]
# Out:
# 0
You can display the matrix
display(A)
FunctionMatrix(m, n, Lambda((i, j), 0))
# Out:
# FunctionMatrix(m, n, Lambda((i, j), 0))
but of course if you try
A.as_explicit()
# Out:
# . . .
# ValueError: Matrix with symbolic shape cannot be represented explicitly.
whereas with non-symbolic values for the numbers of rows/columns you can visualize the matrix:
FunctionMatrix(3, 4, Lambda((i, j), 0)).as_explicit()
# Out:
# Matrix([
# [0, 0, 0, 0],
# [0, 0, 0, 0],
# [0, 0, 0, 0]])
Something like Matrix.zeros won't help because symbols as numbers of rows/columns are not allowed:
from sympy import Matrix, symbols
m, n = symbols('m n', integer=True)
A = Matrix.zeros(m, n) # ❌ gives an error if m,n are symbols
# Out:
# . . .
# TypeError: cannot determine truth value of Relational
m, n = 3, 4 # m and n are not symbols anymore
A = Matrix.zeros(m, n) # ✅ this is ok