erf(x )
=
2
✜
¶
0
x
e
−
t2 dt
For x > -3 and x < 3, the error function looks like this:
erf(x) approaches 1 and -1 for large x and large -x, respectively. Also note that erf(-x) = -erf(x). The
complimentary error function erfc(x) is defined as
erfc(x) = 1 - erfc(x)
This function calculates erf(x) for all real x.
erf(x)
func
©(x) error function of x
©dburkett@infinet.com
1
3 nov 99
©x real and -
1
E-999 <= x <=
1
E999
©Error +/-
1
E-
1
4.
when(x<0,
⁻
erf(
⁻
x),when(x<.84375,x+x*polyeval({
⁻
2.37630
1
6656650
⁻
5,
⁻
5.77027029648
94
⁻
3,
⁻
2.848
1
749575599
⁻
2,
⁻
3.25042
1
07247
⁻1
,
1
.28379
1
670955
1⁻1
},x*x)/polyeval({
⁻
3.9602282787754
⁻
6,
1
.3249473800432
⁻
4,5.08
1
30628
1
8758
⁻
3,6.5022249988767
⁻
2,3.97
9
1
7223959
1
6
⁻1
,
1
},x*x),when(x<
1
.25,.8450629
11
5
1
047+polyeval({
⁻
2.
1
663755948688
⁻
3
,3.54783043256
1
8
⁻
2,
⁻1
.
1
08946942824
⁻1
,3.
1
83466
1
990
11
6
⁻1
,
⁻
3.722078760357
⁻1
,4.
1
4856
11
868375
⁻1
,
⁻
2.362
11
85607527
⁻
3},x-
1
)/polyeval({
1
.
1
984499846799
⁻
2,
1
.3637083
9
1
2029
⁻
2,
1
.26
1
7
1
2
1
980876
⁻1
,7.
1
8286544
1
4
1
96
⁻
2,5.403979
1
7702
1
7
⁻1
,
1
.06420880400
84
⁻1
,
1
},x-
1
),when(x<2.857
1
42857
1
429,
1
-
1
/x*
ℯ
^(
⁻
x*x-.5625+polyeval({
⁻
9.8
1
4329344
1
69
1
,
⁻
8.
1
287435506307
1
,
⁻1
.846050929067
1
2,
⁻1
.6239666946257
2,
⁻
6.2375332450326
1
,
⁻1
.05586262253234
1
,
⁻
6.93858572707
1
8
⁻1
,
⁻
9.864944034847
1⁻
3},
1
/x^2)/polyeval({
⁻
6
.04244
1
52
1
4858
⁻
2,6.5702497703
1
93,
1
.0863500554
1
78
2,4.29008
1
4002757
2,6.4538727
1
73327
2,4.3456587747523
2,
1
.37657754
1
4352
2,
1
.965
1
27
1
667439
1
,
1
},
1
/x^2)),when(x<
5.5
1
8,
1
-
1
/x*
ℯ
^(
⁻
x*x-.5625+polyeval({
⁻
4.835
1
9
1
9
1
60865
2,
⁻1
.025095
1
3
1
6
111
3,
⁻
6.375
6644336839
2,
⁻1
.6063638485582
2,
⁻1
.77579549
1
7755
1
,
⁻
7.9928323768052
⁻1
,
⁻
9.864942
924700
1⁻
3},
1
/x^2)/polyeval({
⁻
2.2440952446586
1
,4.7452854
1
20696
2,2.553050406433
2
3,3.
1
998582
1
95086
3,
1
.5367295860844
3,3.257925
1
299657
2,3.0338060743482
1
,
1
},
1
/x^2)),
1
)))))
Endfunc
This function is accurate to full machine precision for all arguments. The algorithm is taken from
module ERF in the FDLIBM package. This link provides the algorithm details:
http://gams.nist.gov/serve.cgi/Module/FDLIBM/ERF/13299/
6 - 65
Summary of Contents for TI-92+
Page 52: ...Component side of PCB GraphLink I O connector detail 1 41...
Page 53: ...LCD connector detail PCB switch side 1 42...
Page 54: ...Key pad sheet contact side Key pad sheet key side 1 43...
Page 55: ...Key cap detail 1 44...
Page 57: ...Component side of PCB with shield removed A detail view of the intergrated circuits 1 46...
Page 410: ...void extensionroutine2 void Credit to Bhuvanesh Bhatt 10 4...
