Show Menu
Cheatography
  1. Home >
  2. Cheat Sheets >

RenderScript Math Functions Cheat Sheet (DRAFT) by

Overview of RenderScript intrinsic math functions

This is a draft cheat sheet. It is a work in progress and is not finished yet.

Intrinsic math functions

Render­Script provides a lot of intrinsic math functions, which are optimized using NEON instru­ctions (on ARMv7 devices). These functions should be used instead of writing pure calcul­ations when possible.

For example, 
a * b + c
should be translated into 
fma(a, b, c)
.

Note: functions' arguments are assumed to be of type that matches function's return one. Otherwise, different type is specified.
Following functions have been taken directly from Android source code.

Float functions (part 1)

Definition
Descri­ption
 
acos(v)
Inverse cosine
 
acosh(v)
Inverse hyperbolic cosine
 
acospi(v)
Inverse cosine divided by pi
 
asin(v)
Inverse sine
 
asinh(v)
Inverse hyperbolic sine
 
asinpi(v)
Inverse sine divided by pi
 
atan(v)
Inverse tangent
 
atan2(num, den)
Inverse tangent of a ratio
 
atan2p­i(num, den)
Inverse tangent of a ratio, divided by pi
 
atanh(v)
Inverse hyperbolic tangent
 
atanpi(v)
Inverse tangent divided by pi
 
cbrt(v)
Cube root
 
ceil(v)
Smallest integer not less than a value
 
clamp(­value, min, max)
Restrain a value to a range
 
copysi­gn(­mag­nitude, sign)
Copies the sign of a number to another. 
copysi­gn(­4.0f, -2.7f)
returns 
-4.0f
 
cos(v)
Cosine
 
cosh(v)
Hypebolic cosine
 
cospi(v)
Cosine of a number multiplied by pi
 
degrees(v)
Converts radians into degrees
 
erf(v)
Mathem­atical error function
 
erfc(v)
Mathem­atical comple­mentary error function
 
exp(v)
e raised to a number
 
exp10(v)
10 raised to a number
 
exp2(v)
2 raised to a number
 
expm1(v)
e raised to a number minus one: 
(e ^ v) - 1
 
fabs(v)
Absolute value of a float
 
fdim(a, b)
Positive difference between two values. If 
a > b
, returns 
(a - b)
otherwise returns 
0f
 
floor(v)
Smallest integer not greater than a value/ 
(multi­pli­cand1 * multip­lic­and2) + offset
 
fma(mul1, mul2, offset)
Multiply and add
 
fmax(a, b)
Maximum of two floats
 
fmin(a, b)
Minimum of two floats
 
fmod(num, den)
Modulo
 
fract(v, float* floor)
Positive fractional part. 
fract(­2.3f, &val)
returns 
0.3f
and sets val to 
2.f
 
frexp(v, int* exponent)
Binary mantissa and exponent
 
half_r­ecip(v)
Reciprocal computed to 16 bit precision
 
half_r­sqrt(v)
Reciprocal of a square root computed to 16 bit precision
 
half_s­qrt(v)
Square root computed to 16 bit precision
 

Predefined constants

Type
Name
Value
 
float
 
M_PI
 
3.1415­92... (pi)
 
float
 
M_PI_2
 
M_PI / 2
 
float
 
M_PI_4
 
M_PI / 4
 
float
 
M_1_PI
 
1 / M_PI
 
float
 
M_2_PI
 
2 / M_PI
 
float
 
M_2_SQRTPI
 
2 / sqrt(M_PI)
 
float
 
M_SQRT2
 
sqrt(2)
 
float
 
M_E
 
2.7182­81... (e)
 
float
 
M_LN10
 
log_e(10)
 
float
 
M_LN2
 
log_e(2)
 
float
 
M_LOG10E
 
log_10­(M_E)
 
float
 
M_LOG2E
 
log_2(M_E)
 
float
 
M_SQRT1_2
 
1 / sqrt(2)

Float functions (part 2)

Definition
Descri­ption
 
hypot(a, b)
Hypotenuse
 
ldexp(­man­tissa, int exponent)
Creates a floating point from mantissa and exponent. 
mantissa * 2 ^ exponent
 
lgamma(v)
Natural logarithm of the gamma function
 
lgamma(v, int* sign_o­f_g­amma)
Natural logarithm of the gamma function. If 
sign_o­f_gamma
is not null, 
*sign_­of_­gamma
will be set to 
-1.f
if the gamma of 
v
is negative, otherwise to 
1.f
 
log(v)
Natural logarithm
 
log10(v)
Base 10 logarithm
 
log1p(v)
Natural logarithm of a value plus 1
 
log2(v)
Base 2 logarithm
 
logb(v)
Base two exponent. 
logb(8.5f)
returns 
3.f
 
mad(mul1, mul2, offset)
Multiply and add
 
max(a, b)
Maximum
 
min(a, b)
Minimum
 
modf(v, float* integr­al_­part)
Integral and fractional compon­ents. 
modf(-­3.72f)
will return 
0.72f
and 
integr­al_part
will be set to 
-3.f
 
nan(uint v)
Returns NaN
 
nextaf­ter(v, target)
Next repres­entable floating point number from v towards target
 
pow(base, exp)
Base raised to an exponent
 
pown(base, int expt)
Base raised to an integer exponent
 
powr(base, exp)
Positive base raised to an exponent
 
radians(v)
Converts degrees into radians
 
remain­der­(num, den)
Remainder of a division
 
remquo­(num, int* quotient)
Remainder and quotient of a division
 
rint(v)
Round to even
 
rootn(v, int n)
Nth root
 
round(v)
Round away from zero
 
rsqrt(v)
Reciprocal of a square root
 
sign(v)
Sign of a value
 
sin(v)
Sine
 
sincos(v, float* cos)
Sine and cosine
 
sinh(v)
Hyperbolic sine
 
sinpi(v)
Sine of a number multiplied by pi
 
sqrt(v)
Square root
 
step(edge, v)
Returns 
0.f
if 
v < edge
, 
1.f
otherwise
 
tan(v)
Tangent
 
tanh(v)
Hyperbolic tangent
 
tanpi(v)
Tangent of a number multiplied by pi
 
tgamma(v)
Gamma function
 
trunc(v)
Truncates a floating point
 

Integer functions (return int)

Definition
Descri­ption
 
abs(v)
Absolute value of an integer
 
clamp(­value, min, max)
Restrain a value to a range (min API 19)
 
clz(value)
Number of leading 0 bits
 
ilogb(­float v)
Base two exponent
 
max(a, b)
Maximum value of two arguments
 
min(a, b)
Minimum value of two arguments
 
rsRand­(max)
Pseudo­-random number
 
rsRand­(min, max)
Pseudo­-random number

Approx­imate float functions (API >= 21)

Following functions have stricter limits than precise ones. Please refer to specs to see them.
Definition
Descri­ption
 
native­_lo­g2(v)
Approx­imate base 2 logarithm (API 18)
 
native­_po­wr(­base, exp)
Approx­imate positive base raised to an exponent (API 18)
 
native­_ac­os(v)
Approx­imate inverse cosine
 
native­_ac­osh(v)
Approx­imate inverse hyperbolic cosine
 
native­_ac­ospi(v)
Approx­imate inverse cosine divided by pi
 
native­_as­in(v)
Approx­imate inverse sine
 
native­_as­inh(v)
Approx­imate inverse hyperbolic sine
 
native­_as­inpi(v)
Approx­imate inverse sine divided by pi
 
native­_at­an(v)
Approx­imate inverse tangent
 
native­_at­an2­(num, den)
Approx­imate inverse tangent of a ratio
 
native­_at­an2­pi(num, den)
Approx­imate inverse tangent of a ratio, divided by pi
 
native­_at­anh(v)
Approx­imate inverse hyperbolic tangent
 
native­_at­anpi(v)
Approx­imate inverse tangent divided by pi
 
native­_cb­rt(v)
Approx­imate cube root
 
native­_cos(v)
Approx­imate cosine
 
native­_co­sh(v)
Approx­imate hypebolic cosine
 
native­_co­spi(v)
Approx­imate cosine of a number multiplied by pi
 
native­_di­vid­e(left, right)
Approx­imate division
 
native­_exp(v)
Approx­imate e raised to a number
 
native­_ex­p10(v)
Approx­imate 10 raised to a number
 
native­_ex­p2(v)
Approx­imate 2 raised to a number
 
native­_ex­pm1(v)
Approx­imate e raised to a number minus one
 
native­_hy­pot(a, b)
Approx­imate hypotenuse
 
native­_log(v)
Approx­imate natural logarithm
 
native­_lo­g10(v)
Approx­imate base 10 logarithm
 
native­_lo­g1p(v)
Approx­imate natural logarithm of a value plus 1
 
native­_re­cip(v)
Approx­imate reciprocal
 
native­_ro­otn(v, int n)
Approx­imate nth root
 
native­_rs­qrt(v)
Approx­imate reciprocal of a square root
 
native­_sin(v)
Approx­imate sine
 
native­_si­ncos(v, float* cos);
Approx­imate sine and cosine
 
native­_si­nh(v)
Approx­imate hyperbolic sine
 
native­_si­npi(v)
Approx­imate sine of a number multiplied by pi
 
native­_sq­rt(v)
Approx­imate square root
 
native­_tan(v)
Approx­imate tangent
 
native­_ta­nh(v)
Approx­imate hyperbolic tangent
 
native­_ta­npi(v)
Approx­imate tangent of a number multiplied by pi