Page 58 - Compact Numerical Methods For Computers

TABLE 3.2. Solutions for various principal-component regressions using the data in table 3.1.

Tolerance
for zero R* x c o n s t a n t X nitrogen x phosphate x c o t a s h x petroleum

(a) 0 0·972586 207·782 –0·046191 1·0194 –0·15983 –0·290373
(0·958879)
(b) 1 0·969348 4·33368E–3 –5·85314E–2 1·1757 –0·252296 0·699621
(0·959131)
(c) 22 0·959506 5·14267E–3 4·34851E–2 0·392026 –6·93389E–2 1·0115
(0·951407)
(d) 40 0·93839 2·54597E–3 –0·15299 0·300127 0·469294 0·528881
(0·932789)
Regression involving a constant and index numbers for phosphate

(e) 0 0·968104 179·375 — 0·518966 —
(0·965204)
( f ) 1 0·273448 2·24851E–3 0·945525 —

The values in parentheses below each R’ are the corrected statistic given by formula (3.52).