#  x05.txt
#
#  Reference:
#
#    Helmut Spaeth,
#    Mathematical Algorithms for Linear Regression,
#    Academic Press, 1991, page 305,
#    ISBN 0-12-656460-4.
#
#    Helmut Spaeth,
#    Clusterwise Linear Least Squares Versus Least Absolute Deviation
#      Regression:
#    A Numerical Comparions for a Case Study,
#    In: G Gaul, M Schader, editors,
#    Classification as a Tool of Research,
#    North Holland, 1986.
#
#  Discussion:
#
#    Mail order houses issue supplementary lists with varying print runs
#    and different numbers of pages.  Each supplementary list
#    produces a certain number of orders.  
#
#    There are 38 rows of data.  The data columns include:
#
#      I,  the index;
#      A0, 1;
#      A1. the size of the print run, in thousands;
#      A2, the number of pages in the list;
#      B,  the number of orders, in thousands.
#
#    We seek a model of the form:
#
#      B = A0 * X0 + A1 * X1 + A2 * X2.
#
5 columns
38 rows
Index
One
Print Runs (x1000)
Page numbers
Orders (x1000)
 1  1  2800   22.   437.
 2  1  2670   14.   204.
 3  1  2800   37.   725.
 4  1  2784   15.   279.
 5  1  2800   38.   474.
 6  1  2620  172.  1587.
 7  1  2620  249.  2630.
 8  1  2470   84.   798.
 9  1  2620  242.  2509.
10  1  2475  100.  1192.
11  1  2620  114.   882.
12  1  2620   37.   511.
13  1  2448   96.   896.
14  1  2648  116.  1297.
15  1  2525   94.   857.
16  1  1000   47.   388.
17  1   980   48.   462.
18  1  1000   15.    67.
19  1  1112   45.   326.
20  1  1000   23.   145.
21  1  1000   44.   298.
22  1  2188   23.   179.
23  1  1028   47.   289.
24  1  2200   31.   200.
25  1  1000   48.   461.
26  1   980   48.   223.
27  1   728   47.   235.
28  1  2510   26.   235.
29  1  1500   58.   594.
30  1  2500  128.  1800.
31  1  2620  120.  1457.
32  1  2528  120.  1710.
33  1  2630  121.  1715.
34  1  2550  122.  1615.
35  1  1150   61.   196.
36  1  1150   50.   309.
37  1  1150   50.   263.
38  1  1147   60.   332.