I run a I-O table. I try to estimate the Leontied inverse and finally got to know that my CILQ (FLQ) was wrong. I checked for infinities and this is the source of the problem. I mark the problematic line with Bold. Lines in Italics are the results of the check (to save time).
EmpRatio=EmpRg./EmpNt;
Matrix EmpRg is finite
Matrix EmpNt is finite
Matrix EmpRatio is finite
C1=reshape(EmpRatio,m,m);
Matrix C1 is finite
CILQ0=(EmpRatio./C1)* 0.34181;
CILQ0 has inifinities or NaN's