Question: Non-linear least square implicit function

I have a dataset:

NMP:=<0.530,0.555,0.572,0.592>:
ETOH:=<0.136,0.153,0.163,0.170>:

For these four data points [NMP,ETOH] I want to find the least square function in form of:

ln(ETOH/(1-ETOH-NMP))=a*ln(NMP/(1-ETOH-NMP))+b

also I need to find appropriate a,b constant values.

This function is implicite so I cannot use with(Statistics):NonlinearFit.

Can you help me how to determine a,b constants?

Please Wait...