Previous Up Next

7.3.11  Error function

The error function erf is defined by:

  erf(x)=
2
√
π
∫
x


0
e−t2dt,

where the constant 2/√π is chosen so that erf(+∞)=1 and erf(−∞)=−1, since

  ∫
+∞


0
e−t2dt=
√
π
2
.

The erf command computes the error function.

Examples

erf(1)
     
erf⎛
⎝
1⎞
⎠
          
erf(1.0)
     
0.84270079295           
erf(1/(sqrt(2)))*1/2+0.5
     
0.841344746069           
Remark.

The relation between erf and normal_cdf (see Section 20.4.7) is:

  normal_cdf(x)=
1
2
+
1
2
erf⎛
⎜
⎜
⎜
⎜
⎝
x
√
2
⎞
⎟
⎟
⎟
⎟
⎠

Indeed, making the change of variable t=u√2 in normal_cdf(x)=1/2+1/√2π∫0xe−t2/2dt yields:

  normal_cdf(x)=
1
2
+
1
√
π
∫
x
√
2


0
e−u2du=
1
2
+
1
2
erf⎛
⎜
⎜
⎜
⎜
⎝
x
√
2
⎞
⎟
⎟
⎟
⎟
⎠
.

Previous Up Next