cook_function

Convert a single function call into a flattened variable symbol.

Transforms a function expression with arguments into a single symbolic variable whose name encodes both the function name and its arguments. This is the atomic version of cook_indices(), operating on individual function calls rather than complete expressions.

Function Cooking Transform:

• Simple function: f(a, b) → f_a_b
• Nested arguments: tensor(rep(mu)) → tensor_rep_mu
• Multiple arguments: gamma(alpha, beta, mu) → gamma_alpha_beta_mu
• Complex names: my_function(x, y) → my_function_x_y

Constraints:

• Input must be a single function call (not sum, product, etc.)
• Arguments must be cookable (symbols, numbers, simple functions)
• Cannot cook expressions containing polynomials or complex structures

import symbolica as sp
from symbolica.community.idenso import cook_function

# Simple function cooking
f = sp.S('f')
a, b = sp.S('a','b')

cooked = cook_function(f(a, b))
print(cooked)  # Outputs: f_a_b