v/vlib/math/erf.v

373 lines
11 KiB
V

module math
/*
* x
* 2 |\
* erf(x) = --------- | exp(-t*t)dt
* sqrt(pi) \|
* 0
*
* erfc(x) = 1-erf(x)
* Note that
* erf(-x) = -erf(x)
* erfc(-x) = 2 - erfc(x)
*
* Method:
* 1. For |x| in [0, 0.84375]
* erf(x) = x + x*R(x**2)
* erfc(x) = 1 - erf(x) if x in [-.84375,0.25]
* = 0.5 + ((0.5-x)-x*R) if x in [0.25,0.84375]
* where R = P/Q where P is an odd poly of degree 8 and
* Q is an odd poly of degree 10.
* -57.90
* | R - (erf(x)-x)/x | <= 2
*
*
* Remark. The formula is derived by noting
* erf(x) = (2/sqrt(pi))*(x - x**3/3 + x**5/10 - x**7/42 + ....)
* and that
* 2/sqrt(pi) = 1.128379167095512573896158903121545171688
* is close to one. The interval is chosen because the fix
* point of erf(x) is near 0.6174 (i.e., erf(x)=x when x is
* near 0.6174), and by some experiment, 0.84375 is chosen to
* guarantee the error is less than one ulp for erf.
*
* 2. For |x| in [0.84375,1.25], let s_ = |x| - 1, and
* c = 0.84506291151 rounded to single (24 bits)
* erf(x) = sign(x) * (c + P1(s_)/Q1(s_))
* erfc(x) = (1-c) - P1(s_)/Q1(s_) if x > 0
* 1+(c+P1(s_)/Q1(s_)) if x < 0
* |P1/Q1 - (erf(|x|)-c)| <= 2**-59.06
* Remark: here we use the taylor series expansion at x=1.
* erf(1+s_) = erf(1) + s_*Poly(s_)
* = 0.845.. + P1(s_)/Q1(s_)
* That is, we use rational approximation to approximate
* erf(1+s_) - (c = (single)0.84506291151)
* Note that |P1/Q1|< 0.078 for x in [0.84375,1.25]
* where
* P1(s_) = degree 6 poly in s_
* Q1(s_) = degree 6 poly in s_
*
* 3. For x in [1.25,1/0.35(~2.857143)],
* erfc(x) = (1/x)*exp(-x*x-0.5625+R1/s1)
* erf(x) = 1 - erfc(x)
* where
* R1(z) = degree 7 poly in z, (z=1/x**2)
* s1(z) = degree 8 poly in z
*
* 4. For x in [1/0.35,28]
* erfc(x) = (1/x)*exp(-x*x-0.5625+R2/s2) if x > 0
* = 2.0 - (1/x)*exp(-x*x-0.5625+R2/s2) if -6<x<0
* = 2.0 - tiny (if x <= -6)
* erf(x) = sign(x)*(1.0 - erfc(x)) if x < 6, else
* erf(x) = sign(x)*(1.0 - tiny)
* where
* R2(z) = degree 6 poly in z, (z=1/x**2)
* s2(z) = degree 7 poly in z
*
* Note1:
* To compute exp(-x*x-0.5625+R/s), let s_ be a single
* precision number and s_ := x; then
* -x*x = -s_*s_ + (s_-x)*(s_+x)
* exp(-x*x-0.5626+R/s) =
* exp(-s_*s_-0.5625)*exp((s_-x)*(s_+x)+R/s);
* Note2:
* Here 4 and 5 make use of the asymptotic series
* exp(-x*x)
* erfc(x) ~ ---------- * ( 1 + Poly(1/x**2) )
* x*sqrt(pi)
* We use rational approximation to approximate
* g(s_)=f(1/x**2) = log(erfc(x)*x) - x*x + 0.5625
* Here is the error bound for R1/s1 and R2/s2
* |R1/s1 - f(x)| < 2**(-62.57)
* |R2/s2 - f(x)| < 2**(-61.52)
*
* 5. For inf > x >= 28
* erf(x) = sign(x) *(1 - tiny) (raise inexact)
* erfc(x) = tiny*tiny (raise underflow) if x > 0
* = 2 - tiny if x<0
*
* 7. special case:
* erf(0) = 0, erf(inf) = 1, erf(-inf) = -1,
* erfc(0) = 1, erfc(inf) = 0, erfc(-inf) = 2,
* erfc/erf(nan) is nan
*/
const erx = 8.45062911510467529297e-01 // 0x3FEB0AC160000000
// Coefficients for approximation to erf in [0, 0.84375]
const efx = 1.28379167095512586316e-01 // 0x3FC06EBA8214DB69
const efx8 = 1.02703333676410069053e+00 // 0x3FF06EBA8214DB69
const pp0 = 1.28379167095512558561e-01 // 0x3FC06EBA8214DB68
const pp1 = -3.25042107247001499370e-01 // 0xBFD4CD7D691CB913
const pp2 = -2.84817495755985104766e-02 // 0xBF9D2A51DBD7194F
const pp3 = -5.77027029648944159157e-03 // 0xBF77A291236668E4
const pp4 = -2.37630166566501626084e-05 // 0xBEF8EAD6120016AC
const qq1 = 3.97917223959155352819e-01 // 0x3FD97779CDDADC09
const qq2 = 6.50222499887672944485e-02 // 0x3FB0A54C5536CEBA
const qq3 = 5.08130628187576562776e-03 // 0x3F74D022C4D36B0F
const qq4 = 1.32494738004321644526e-04 // 0x3F215DC9221C1A10
const qq5 = -3.96022827877536812320e-06 // 0xBED09C4342A26120
// Coefficients for approximation to erf in [0.84375, 1.25]
const pa0 = -2.36211856075265944077e-03 // 0xBF6359B8BEF77538
const pa1 = 4.14856118683748331666e-01 // 0x3FDA8D00AD92B34D
const pa2 = -3.72207876035701323847e-01 // 0xBFD7D240FBB8C3F1
const pa3 = 3.18346619901161753674e-01 // 0x3FD45FCA805120E4
const pa4 = -1.10894694282396677476e-01 // 0xBFBC63983D3E28EC
const pa5 = 3.54783043256182359371e-02 // 0x3FA22A36599795EB
const pa6 = -2.16637559486879084300e-03 // 0xBF61BF380A96073F
const qa1 = 1.06420880400844228286e-01 // 0x3FBB3E6618EEE323
const qa2 = 5.40397917702171048937e-01 // 0x3FE14AF092EB6F33
const qa3 = 7.18286544141962662868e-02 // 0x3FB2635CD99FE9A7
const qa4 = 1.26171219808761642112e-01 // 0x3FC02660E763351F
const qa5 = 1.36370839120290507362e-02 // 0x3F8BEDC26B51DD1C
const qa6 = 1.19844998467991074170e-02 // 0x3F888B545735151D
// Coefficients for approximation to erfc in [1.25, 1/0.35]
const ra0 = -9.86494403484714822705e-03 // 0xBF843412600D6435
const ra1 = -6.93858572707181764372e-01 // 0xBFE63416E4BA7360
const ra2 = -1.05586262253232909814e+01 // 0xC0251E0441B0E726
const ra3 = -6.23753324503260060396e+01 // 0xC04F300AE4CBA38D
const ra4 = -1.62396669462573470355e+02 // 0xC0644CB184282266
const ra5 = -1.84605092906711035994e+02 // 0xC067135CEBCCABB2
const ra6 = -8.12874355063065934246e+01 // 0xC054526557E4D2F2
const ra7 = -9.81432934416914548592e+00 // 0xC023A0EFC69AC25C
const sa1 = 1.96512716674392571292e+01 // 0x4033A6B9BD707687
const sa2 = 1.37657754143519042600e+02 // 0x4061350C526AE721
const sa3 = 4.34565877475229228821e+02 // 0x407B290DD58A1A71
const sa4 = 6.45387271733267880336e+02 // 0x40842B1921EC2868
const sa5 = 4.29008140027567833386e+02 // 0x407AD02157700314
const sa6 = 1.08635005541779435134e+02 // 0x405B28A3EE48AE2C
const sa7 = 6.57024977031928170135e+00 // 0x401A47EF8E484A93
const sa8 = -6.04244152148580987438e-02 // 0xBFAEEFF2EE749A62
// Coefficients for approximation to erfc in [1/.35, 28]
const rb0 = -9.86494292470009928597e-03 // 0xBF84341239E86F4A
const rb1 = -7.99283237680523006574e-01 // 0xBFE993BA70C285DE
const rb2 = -1.77579549177547519889e+01 // 0xC031C209555F995A
const rb3 = -1.60636384855821916062e+02 // 0xC064145D43C5ED98
const rb4 = -6.37566443368389627722e+02 // 0xC083EC881375F228
const rb5 = -1.02509513161107724954e+03 // 0xC09004616A2E5992
const rb6 = -4.83519191608651397019e+02 // 0xC07E384E9BDC383F
const sb1 = 3.03380607434824582924e+01 // 0x403E568B261D5190
const sb2 = 3.25792512996573918826e+02 // 0x40745CAE221B9F0A
const sb3 = 1.53672958608443695994e+03 // 0x409802EB189D5118
const sb4 = 3.19985821950859553908e+03 // 0x40A8FFB7688C246A
const sb5 = 2.55305040643316442583e+03 // 0x40A3F219CEDF3BE6
const sb6 = 4.74528541206955367215e+02 // 0x407DA874E79FE763
const sb7 = -2.24409524465858183362e+01
// erf returns the error function of x.
//
// special cases are:
// erf(+inf) = 1
// erf(-inf) = -1
// erf(nan) = nan
pub fn erf(a f64) f64 {
mut x := a
very_tiny := 2.848094538889218e-306 // 0x0080000000000000
small_ := 1.0 / f64(u64(1) << 28) // 2**-28
if is_nan(x) {
return nan()
}
if is_inf(x, 1) {
return 1.0
}
if is_inf(x, -1) {
return f64(-1)
}
mut sign := false
if x < 0 {
x = -x
sign = true
}
if x < 0.84375 { // |x| < 0.84375
mut temp := 0.0
if x < small_ { // |x| < 2**-28
if x < very_tiny {
temp = 0.125 * (8.0 * x + efx8 * x) // avoid underflow
} else {
temp = x + efx * x
}
} else {
z := x * x
r := pp0 + z * (pp1 + z * (pp2 + z * (pp3 + z * pp4)))
s_ := 1.0 + z * (qq1 + z * (qq2 + z * (qq3 + z * (qq4 + z * qq5))))
y := r / s_
temp = x + x * y
}
if sign {
return -temp
}
return temp
}
if x < 1.25 { // 0.84375 <= |x| < 1.25
s_ := x - 1
p := pa0 + s_ * (pa1 + s_ * (pa2 + s_ * (pa3 + s_ * (pa4 + s_ * (pa5 + s_ * pa6)))))
q := 1.0 + s_ * (qa1 + s_ * (qa2 + s_ * (qa3 + s_ * (qa4 + s_ * (qa5 + s_ * qa6)))))
if sign {
return -erx - p / q
}
return erx + p / q
}
if x >= 6 { // inf > |x| >= 6
if sign {
return -1
}
return 1.0
}
s_ := 1.0 / (x * x)
mut r := 0.0
mut s := 0.0
if x < 1.0 / 0.35 { // |x| < 1 / 0.35 ~ 2.857143
r = ra0 + s_ * (ra1 + s_ * (ra2 + s_ * (ra3 + s_ * (ra4 + s_ * (ra5 + s_ * (ra6 +
s_ * ra7))))))
s = 1.0 + s_ * (sa1 + s_ * (sa2 + s_ * (sa3 + s_ * (sa4 + s_ * (sa5 + s_ * (sa6 +
s_ * (sa7 + s_ * sa8)))))))
} else { // |x| >= 1 / 0.35 ~ 2.857143
r = rb0 + s_ * (rb1 + s_ * (rb2 + s_ * (rb3 + s_ * (rb4 + s_ * (rb5 + s_ * rb6)))))
s = 1.0 + s_ * (sb1 + s_ * (sb2 + s_ * (sb3 + s_ * (sb4 + s_ * (sb5 + s_ * (sb6 +
s_ * sb7))))))
}
z := f64_from_bits(f64_bits(x) & 0xffffffff00000000) // pseudo-single (20-bit) precision x
r_ := exp(-z * z - 0.5625) * exp((z - x) * (z + x) + r / s)
if sign {
return r_ / x - 1.0
}
return 1.0 - r_ / x
}
// erfc returns the complementary error function of x.
//
// special cases are:
// erfc(+inf) = 0
// erfc(-inf) = 2
// erfc(nan) = nan
pub fn erfc(a f64) f64 {
mut x := a
tiny := 1.0 / f64(u64(1) << 56) // 2**-56
// special cases
if is_nan(x) {
return nan()
}
if is_inf(x, 1) {
return 0.0
}
if is_inf(x, -1) {
return 2.0
}
mut sign := false
if x < 0 {
x = -x
sign = true
}
if x < 0.84375 { // |x| < 0.84375
mut temp := 0.0
if x < tiny { // |x| < 2**-56
temp = x
} else {
z := x * x
r := pp0 + z * (pp1 + z * (pp2 + z * (pp3 + z * pp4)))
s_ := 1.0 + z * (qq1 + z * (qq2 + z * (qq3 + z * (qq4 + z * qq5))))
y := r / s_
if x < 0.25 { // |x| < 1.0/4
temp = x + x * y
} else {
temp = 0.5 + (x * y + (x - 0.5))
}
}
if sign {
return 1.0 + temp
}
return 1.0 - temp
}
if x < 1.25 { // 0.84375 <= |x| < 1.25
s_ := x - 1
p := pa0 + s_ * (pa1 + s_ * (pa2 + s_ * (pa3 + s_ * (pa4 + s_ * (pa5 + s_ * pa6)))))
q := 1.0 + s_ * (qa1 + s_ * (qa2 + s_ * (qa3 + s_ * (qa4 + s_ * (qa5 + s_ * qa6)))))
if sign {
return 1.0 + erx + p / q
}
return 1.0 - erx - p / q
}
if x < 28 { // |x| < 28
s_ := 1.0 / (x * x)
mut r := 0.0
mut s := 0.0
if x < 1.0 / 0.35 { // |x| < 1 / 0.35 ~ 2.857143
r = ra0 + s_ * (ra1 + s_ * (ra2 + s_ * (ra3 + s_ * (ra4 + s_ * (ra5 + s_ * (ra6 +
s_ * ra7))))))
s = 1.0 + s_ * (sa1 + s_ * (sa2 + s_ * (sa3 + s_ * (sa4 + s_ * (sa5 + s_ * (sa6 +
s_ * (sa7 + s_ * sa8)))))))
} else { // |x| >= 1 / 0.35 ~ 2.857143
if sign && x > 6 {
return 2.0 // x < -6
}
r = rb0 + s_ * (rb1 + s_ * (rb2 + s_ * (rb3 + s_ * (rb4 + s_ * (rb5 + s_ * rb6)))))
s = 1.0 + s_ * (sb1 + s_ * (sb2 + s_ * (sb3 + s_ * (sb4 + s_ * (sb5 + s_ * (sb6 +
s_ * sb7))))))
}
z := f64_from_bits(f64_bits(x) & 0xffffffff00000000) // pseudo-single (20-bit) precision x
r_ := exp(-z * z - 0.5625) * exp((z - x) * (z + x) + r / s)
if sign {
return 2.0 - r_ / x
}
return r_ / x
}
if sign {
return 2.0
}
return 0.0
}