- #1
RedX
- 970
- 3
The integral of ln(1-x) is -(1-x)ln(1-x)-x, when 0<=x<=1.
So for example:
[tex]\int_{0}^{1} ln (1-x)dx= (-(1-x)ln(1-x)+x)_{0}^{1}=-1 [/tex]
However, going to an online integrating site:
http://integrals.wolfram.com/index.jsp?expr=Log[1-x]&random=false
they give the integral of ln(1-x) as x*(-1 + ln[1 - x]) - ln[-1 + x].
So according to them:
[tex]\int_{0}^{1} ln (1-x)dx=(x*(-1 + ln[1 - x]) - ln[-1 + x])_{0}^{1} =
1(-1+ln(0))-ln(0)-0+ln(-1)=-1+ln(-1) [/tex]
For ln(-1), aren't you supposed to take the value [tex]i\pi[/tex]?
So do online integrators arbitrarily change signs of the argument in ln(), and give you an answer, with the understanding that it's only correct up to an imaginary part that's the result of assigning a sign to the argument of the ln()? Are they careful when you multiply ln()s, like ln(x)*ln(1-x), where multiplying two imaginary numbers gives a real number, so you can't tell if a term is due to assigning the argument of a logarithm a certain sign?
So for example:
[tex]\int_{0}^{1} ln (1-x)dx= (-(1-x)ln(1-x)+x)_{0}^{1}=-1 [/tex]
However, going to an online integrating site:
http://integrals.wolfram.com/index.jsp?expr=Log[1-x]&random=false
they give the integral of ln(1-x) as x*(-1 + ln[1 - x]) - ln[-1 + x].
So according to them:
[tex]\int_{0}^{1} ln (1-x)dx=(x*(-1 + ln[1 - x]) - ln[-1 + x])_{0}^{1} =
1(-1+ln(0))-ln(0)-0+ln(-1)=-1+ln(-1) [/tex]
For ln(-1), aren't you supposed to take the value [tex]i\pi[/tex]?
So do online integrators arbitrarily change signs of the argument in ln(), and give you an answer, with the understanding that it's only correct up to an imaginary part that's the result of assigning a sign to the argument of the ln()? Are they careful when you multiply ln()s, like ln(x)*ln(1-x), where multiplying two imaginary numbers gives a real number, so you can't tell if a term is due to assigning the argument of a logarithm a certain sign?