- daily to monthly
 - daily or monthly to quaterly

Conversion to annual frequency is not yet implemented, but this is
enough to close #1.

Four time aggregation methods are available:

 - 'sum' (aggregation of flow variables),
 - 'end-of-period' (aggregation of stock variables),
 - 'arithmetic-average', (aggregation of rates), and
 - 'geometric-average' (aggregation of ratios)

Examples:

>> a = dseries(rand(2*365+1,1), '2000-01-01')

a is a dseries object:

           | Variable_1
2000-01-01 | 0.26032
2000-01-02 | 0.56927
2000-01-03 | 0.24877
2000-01-04 | 0.3193
2000-01-05 | 0.9108
2000-01-06 | 0.88522
2000-01-07 | 0.79459
2000-01-08 | 0.92581
2000-01-09 | 0.17884
2000-01-10 | 0.51754
           |
2001-12-21 | 0.035016
2001-12-22 | 0.03975
2001-12-23 | 0.98857
2001-12-24 | 0.68618
2001-12-25 | 0.37669
2001-12-26 | 0.50432
2001-12-27 | 0.7635
2001-12-28 | 0.048875
2001-12-29 | 0.72593
2001-12-30 | 0.70133
2001-12-31 | 0.45889

>> dseries2Q(a, 'sum')

ans is a dseries object:

       | Variable_1
2000Q1 | 46.6096
2000Q2 | 45.1456
2000Q3 | 45.0172
2000Q4 | 47.6164
2001Q1 | 49.9093
2001Q2 | 46.2334
2001Q3 | 47.4685
2001Q4 | 47.1611

>> dseries2M(a, 'sum')

ans is a dseries object:

        | Variable_1
2000M1  | 16.7499
2000M2  | 15.4311
2000M3  | 14.4286
2000M4  | 14.3583
2000M5  | 15.7072
2000M6  | 15.0801
2000M7  | 14.2406
2000M8  | 15.9174
2000M9  | 14.8592
2000M10 | 17.5525
2000M11 | 13.3236
2000M12 | 16.7404
2001M1  | 18.9293
2001M2  | 16.1282
2001M3  | 14.8518
2001M4  | 16.9843
2001M5  | 14.5768
2001M6  | 14.6723
2001M7  | 14.9768
2001M8  | 15.3968
2001M9  | 17.0949
2001M10 | 15.0376
2001M11 | 16.6848
2001M12 | 15.4388

>> dseries2Q(dseries2M(a, 'sum'), 'sum')

ans is a dseries object:

       | Variable_1
2000Q1 | 46.6096
2000Q2 | 45.1456
2000Q3 | 45.0172
2000Q4 | 47.6164
2001Q1 | 49.9093
2001Q2 | 46.2334
2001Q3 | 47.4685
2001Q4 | 47.1611

Conversion of daily dates objects to lower frequency.

Frequency is passed as a member of the dates class.

Also removed reference to weekly frequency.

Also added and updated unit tests.