ADDNOISE

Adds noise to model data

Description:

ADDNOISE, adds noise to model data. The noise can be either poissonian or gaussian. If gaussian a constant noise level, described by a given standard deviation, is introduced. If poissonian the data is scaled by a factor to change data values into counts. The counts are then used as estimates of the mean value in that pixel and noise is added on this basis. Note that the poissonian noise is pseudo gaussian and so the count levels in the frame need to be greater than 10.

Each time this program is started a different set of random numbers should be generated.

Usage:

ADDNOISE IN NOISE OUT
 
ADU=? SIGMA=?

  noise

Parameters:

IN = NDF (Read)
The input NDF containing the data to which noise needs to be added.
NOISE = LITERAL (Read)
The noise type to introduce into the data. Either Gaussian or Poissonian, which may be abbreviated to G and P. [P]
ADU = REAL (Read)
The scaling factor to convert the data values in the input NDF to counts for which Poisson statistics are assumed valid. [1.0]
SIGMA = REAL (Read)
The standard deviation of the gaussian noise. [1.0]
OUT = NDF (Write)
The output NDF to contain the data with noise added.

Examples:

ADDNOISE IN=MODEL NOISE=G SIGMA=20 OUT=MODEL_WITH_NOISE
This adds gaussian noise to the input NDF model. The noise has a standard deviation of 20 units.
ADDNOISE IN=MODEL NOISE=P ADU=10 OUT=MODEL_WITH_NOISE
This adds poissonian noise to the input NDF model. The data values are scaled by a factor of 10 before the noise is calculated.