Quantizing Data with quantizer Objects

You construct a quantizer object to specify the quantization parameters to use when you quantize data sets. You can use the quantize function to quantize data according to a quantizer object's specifications.

Once you quantize data with a quantizer object, its state values might change.

The following example shows

• How you use quantize to quantize data

• How quantization affects quantizer object states

• How you reset quantizer object states to their default values using reset

1. Construct an example data set and a quantizer object.

   format long g
   randn('state',0);
   x = randn(100,4);
   q = quantizer([16,14]);

2. Retrieve the values of the maxlog and noverflows states.

   q.maxlog
   
   ans =
   
       -1.79769313486232e+308
   
   q.noverflows
   
   ans =
   
        0

   Note that maxlog is equal to -realmax, which indicates that the quantizer q is in a reset state.

3. Quantize the data set according to the quantizer object's specifications.

   y = quantize(q,x);
   Warning: 15 overflows.

4. Check the values of maxlog and noverflows.

   q.maxlog
   
   ans =
   
             1.99993896484375
   
   q.noverflows
   
   ans =
   
       15

   Note that the maximum logged value was taken after quantization, that is, q.maxlog == max(y).

5. Reset the quantizer states and check them.

   reset(q)
   q.maxlog
   
   ans =
   
       -1.79769313486232e+308
   
   q.noverflows
   
   ans =
   
        0

quantizer Object Properties

Transformations for Quantized Data

© 1984-2012- The MathWorks, Inc.    -   Site Help   -   Patents   -   Trademarks   -   Privacy Policy   -   Preventing Piracy   -   RSS