Hi,
I have published the next post at "Analysis, Topology and Differential Geometry" but got no replays. I hope that here someone will be able to help me.
I'm trying to implement factor analysis but i cannot extract the seccund factor
I have the next correlation matrix:
1 0.9848 0.342 0 0 -0.1736
0.9848 1 0.5 0.1736 0
0.342 0.5 1 0.9397 0.866
0 0.1736 0.9398 1 0.9848
-0.1736 0 0.866 0.9858 1
I'm using simple summation method to extract the first factor loading.
and i get correct results:
0.570683029 0.704580979 0.96681105 0.821383882 0.709563721
but in the second factor i have a problem.
the book say :
make new matrix by doing so;
the entry for the correlation between T1 and T2 will be reduced by an amount to l1*l2
the example in the book is:
1-(0.572)^2, and then 1-(0.572*0.705)
(I'm Having some roundoff errors)
the first two results i get are correct but i dont know about the rest.
im continuing 1-(0.572*0.966).... and i get the next matrix
0.674320881 0.582707593 -0.209742658 -0.468749842 -0.578535974
0.582707593 0.503565644 -0.181196676 -0.40513146 -0.499945101
-0.209742658 -0.181196676 0.065276394 0.145576987 0.179985954
-0.468749842 -0.40513146 0.145676987 0.325328518 0.401975796
-0.578535974 -0.499945101 0.179985954 0.402975796 0.496519325
the sum of the first column is zero and in the book final results of the second factor loading shuld be:
-0.98211 -0.7096 0.2554 0.5707 0.7046
what i am doung wrong?