table 7.6 perfectly reproduced, it was numerical problem as well

table 7.6 has been perfectly reproduced.
It was the same numerical problem as table 7.3.
If we format the data with enough digits after decimal point,
then the calculation result from HoltWintersNew() exactly match the data in excel.

For example, for xhat:
the raw data in excel, the calculated data
43.2367816059 43.2367816059
24.9462523004 24.9462523004
33.3589927902 33.3589927902
38.2858278328 38.2858278328
42.2767977876 42.2767977876
26.8902698851 26.8902698851
40.3465246964 40.3465246964
44.0940612746 44.0940612746
50.0542840557 50.0542840557
28.7378995814 28.7378995814
40.8829047371 40.8829047371
45.6530178688 45.6530178688
46.9656969033 46.9656969033
29.9205332297 29.9205332297
40.2965486069 40.2965486069
48.5483325498 48.5483325498
53.2834063916 53.2834063916
32.6493691133 32.6493691133
42.5041359781 42.5041359781
48.6830373900 48.6830373900
56.6772368740 56.6772368740
35.0606731785 35.0606731785
43.9155877711 43.9155877711
50.5291528992 50.5291528992