Multiple regression: Difference between revisions
→{{header|jq}}
(→{{header|Mathematica}}/{{header|Wolfram Language}}: use built-in LeastSquares) |
|||
Line 1,646:
-143.1620228653037
61.960325442985436</pre>
=={{header|jq}}==
{{trans|Wren}}
{{works with|jq}}
'''Works with gojq, the Go implementation of jq'''
See e.g. https://rosettacode.org/wiki/Gauss-Jordan_matrix_inversion#jq
for a suitable definition of `inverse` as used here.
'''Preliminaries'''
<lang jq>def dot_product(a; b):
reduce range(0;a|length) as $i (0; . + (a[$i] * b[$i]) );
# A and B should both be numeric matrices, A being m by n, and B being n by p.
def multiply(A; B):
(B[0]|length) as $p
| (B|transpose) as $BT
| reduce range(0; A|length) as $i
([];
reduce range(0; $p) as $j
(.;
.[$i][$j] = dot_product( A[$i]; $BT[$j] ) ));</lang>
'''Multiple Regression'''
<lang jq>def multipleRegression(y; x):
(y|transpose) as $cy
| (x|transpose) as $cx
| multiply( multiply( multiply(x;$cx)|inverse; x); $cy)
| transpose[0];
def ys: [
[ [1, 2, 3, 4, 5] ],
[ [3, 4, 5] ],
[ [52.21, 53.12, 54.48, 55.84, 57.20, 58.57, 59.93, 61.29,
63.11, 64.47, 66.28, 68.10, 69.92, 72.19, 74.46] ]
];
def a:
[1.47, 1.50, 1.52, 1.55, 1.57, 1.60, 1.63, 1.65, 1.68, 1.70, 1.73, 1.75, 1.78, 1.80, 1.83];
def xs:[
[ [2, 1, 3, 4, 5] ],
[ [1, 2, 1], [1, 1, 2] ],
[ [range(0;a|length) | 1], a, (a|map(.*.))]
];
range(0; ys|length) as $i
| multipleRegression(ys[$i]; xs[$i])</lang>
{{out}}
<pre>
[0.9818181818181818]
[0.9999999999999996,2.000000000000001]
[128.8128035798277,-143.1620228653037,61.960325442985436]
</pre>
=={{header|Julia}}==
| |||