kernelTan
Compute the tangent of a double-precision floating-point number on
[-π/4, π/4].
Usage
var kernelTan = require( '@stdlib/math/base/special/kernel-tan' );
kernelTan( x, y, k )
Computes the tangent of a double-precision floating-point number on [-π/4, π/4].
var out = kernelTan( 3.141592653589793/4.0, 0.0, 1 );
// returns ~1.0
out = kernelTan( 3.141592653589793/6.0, 0.0, 1 );
// returns ~0.577
out = kernelTan( 0.664, 5.288e-17, 1 );
// returns ~0.783
If k = 1, the function returns tan(x+y). To return the negative inverse -1/tan(x+y), set k = -1.
var out = kernelTan( 3.141592653589793/4.0, 0.0, -1 );
// returns ~-1.0
If either x or y is NaN, the function returns NaN.
var out = kernelTan( NaN, 0.0, 1 );
// returns NaN
out = kernelTan( 3.0, NaN, 1 );
// returns NaN
out = kernelTan( NaN, NaN, 1 );
// returns NaN
Notes
For increased accuracy, the number for which the tangent should be evaluated can be supplied as a double-double number (i.e., a non-evaluated sum of two double-precision floating-point numbers
xandy).As components of a double-double number, the two double-precision floating-point numbers
xandymust satisfywhere
ulpstands for units in the last place.
Examples
var linspace = require( '@stdlib/array/base/linspace' );
var binomial = require( '@stdlib/random/base/binomial' ).factory;
var PI = require( '@stdlib/constants/float64/pi' );
var kernelTan = require( '@stdlib/math/base/special/kernel-tan' );
var x = linspace( -PI/4.0, PI/4.0, 100 );
var rbinom = binomial( 1, 0.5 );
var descr;
var i;
var k;
for ( i = 0; i < x.length; i++ ) {
k = rbinom();
descr = ( k === 1 ) ? 'tan(%d) = %d' : '-1/tan(%d) = %d';
console.log( descr, x[ i ], kernelTan( x[ i ], 0.0, k ) );
}
C APIs
Usage
#include "stdlib/math/base/special/kernel_tan.h"
stdlib_base_kernel_tan( x, y, k)
Computes the tangent of a double-precision floating-point number on [-π/4, π/4].
double out = stdlib_base_kernel_tan( 3.141592653589793/4.0, 0.0, 1 );
// returns ~1.0
out = stdlib_base_kernel_tan( 3.141592653589793/6.0, 0.0, 1 );
// returns ~0.577
The function accepts the following arguments:
- x:
[in] doubleinput value (in radians, assumed to be bounded by~pi/4in magnitude). - y:
[in] doubletail ofx. - k:
[in] int32_tindicates whethertan(x+y)(ifk = 1) or-1/tan(x+y)(ifk = -1) is returned.
double stdlib_base_kernel_tan( const double x, const double y, const int32_t k );
Notes
- For increased accuracy, the number for which the tangent should be evaluated can be supplied as a double-double number (i.e., a non-evaluated sum of two double-precision floating-point numbers
xandy).
Examples
#include "stdlib/math/base/special/kernel_tan.h"
#include <stdio.h>
int main( void ) {
const double x[] = { -0.7853981633974483, -0.6108652381980153, -0.4363323129985824, -0.26179938779914946, -0.08726646259971649, 0.08726646259971649, 0.26179938779914935, 0.43633231299858233, 0.6108652381980153, 0.7853981633974483 };
double out;
int i;
for ( i = 0; i < 10; i++ ) {
out = stdlib_base_kernel_tan( x[ i ], 0.0, 1 );
printf( "tan(%lf) = %lf\n", x[ i ], out );
}
}