Feature Overview

• High parsing performance; Bulk mode offers parallelization of expression evaluation.
• if-then-else operator with lazy evaluation
• Compute multiple comma separated subexpressions in a single evaluation: sin(x), x+y, x*x

• 14 predefined operators.
• binary operators are user definable
• postfix operators are user definable
• infix operators are user definable

• 25 predefined functions.
• with a fixed number of up to five arguments
• with variable number of arguments
• with a single string argument (for database queries)
• Can be extended by a user

• numeric constants
• string constants

• unlimited in number
• definable at parser runtime by the parser: abc=123
• assigning variables in terms of other variables is possible

You can define how a value looks like. You can make muparser to read hex values like a=0x00ff00 or binary values. a=0b1000101. You could also use this feature to Query values from a database.

• Use postfix operators as unit multipliers (3m -> 0.003)
• Built in function for numerical differentiation with respect to a given variable.
• Argument separator, decimal separator, thousands separator can be adjusted to your locale
• CMake based build system
• ISO 14882:2017 compliant code
• DLL version usable from every language able to use function exported in C-style

Built-in functions

The following table gives an overview of the functions supported by the default implementation. It lists the function names, the number of arguments and a brief description.

Built-in binary operators

The following table lists the default binary operators supported by the parser.

Ternary Operators

muparser has built in support for the if then else operator. It uses lazy evaluation in order to make sure only the necessary branch of the expression is evaluated.

Built in Constants

The names of predefined constant names are prefixed with an underscore. The parser has two predefined constants: Pi and Eulers number. The accuracy of the constant definition ultimately depends on the size of the selected value type (float, double, long double).