Curve fitting y=a*log(bx)

Hi

I have a data-set of numbers obtained from experiments. When I plot these numbers the distribution look quite similar to y=a*log(bx)

How can I use the least-squares method (or an other method) to find A and B for a fitting curve?

Thanks

Jarno

[JW]

I made one mistake

The function looks more like y=c-a*log(bX)