Continuous Empirical (Histogram) ) Q5 Z" H' q6 u3 K/ N4 F& j
The input data is preprocessed and grouped into k classes. Each class contains all samples within a given interval. The k intervals are disjoint but not necessarily adjacent. , U2 f4 @, X2 p+ m8 U
0 T+ k5 L9 u/ u' X' _; a
The interval includes the lower bound but not the upper bound. If the left bound is equal to the right bound, the interval degenerates to a point. This special case is of interest if the probability of a single point is nonzero.
8 y3 }+ j. B8 \/ f* `2 S" y) f9 A, j) |9 Z
0 [3 i0 z: _" `$ I( X% e5 W/ T
+ c1 K- l; Z, q. F, V# P0 j/ kThe first column of the frequency table contains the lower bound, the second column contains the lower bound, the third column contains the frequency of the corresponding intervals. The data type of the third column has to be real, time, or integer. The number of values of the input data contained in an interval can be used as frequency, for example.
: f7 g4 I+ J0 u. x2 [: F' W1 A+ ?( k2 b; y9 q7 n; v
Usage: z_cEmp(s,tab);
( U ]& M }; x& |! y% ^/ g0 _* Z- i( y
The expression s (integer) defines the random number stream, tab the frequency table. |
[复制链接]
发表于 2011-1-11 10:17:40
楼主