As answer well by @Keith Thompson, the C function works in radians and so a degrees to radian conversion is needed.
#ifndef M_PI
#define M_PI 3.1415926535897932384626433832795
#endif
float angle_radians = angle_degrees * (float) (M_PI/180.0);
Yet rather that directly scale by pi/180.0, code will get more precise answers, for angles outside a primary range, if code first does range reduction and then scales by pi/180.0. This is because the scaling of pi/180.0 is inexact as machine pi float pi = 3.14159265; is not exactly mathematical π. The radian range reduction, performed by cos(), is in error as the computed radian value given to it is inexact to begin with. With degrees, range reduction can be done exactly.
// Benefit with angles outside -360° to +360°
float angle_degrees_reduce = fmodf(angle_degrees, 360.0f);
float angle_radians = angle_degrees_reduce * M_PI/180.0;
sind() is a sine example that performs even better by reducing to a narrower interval. With angles outside the -45° to +45° range, there is benefit.