Is it possible to use a wider data type than float with expr_eval()?

[dipanjan]

Looks like the return type of expr_eval() is float which is 32 bit wide on my 64 bit system. Is it possible to use double which is of 64 bits while evaluating an expression?

[zserge]

Well, expr.h is a single-header library, so you can probably copy it and find/replace all float with double to make it work.

Alternatively, you may replace float with some expr_num_t where appropriate. It would allow you to redefine expr_num_t as double if needed, or any other implementation like fixed-point C++ class with overloaded operators. I haven't thought that far, but I think it might be convenient in some cases.

If you decide to go with expr_num_t way - may I ask you to do a pull request then?