This utility will use a library GenerationLib.py to generate input data suitable for the cases you want to test. It is called by typing GenerateStreamsGUI
. This displays a GUI in which you can select the appropriate parameters to generate the correct input data files:
You have access to a number of parameters:
Data Type: By default,
cint16
, as this is what you use throughout this tutorial.PLIO Width: By default,
64
, as this is the width which is used in this tutorial.Number of Phases: For Super Sampling Rate Filters.
Number of Streams: For the SSR filters using the 2 streams of the AI Engines.
Number of Samples per Stream per Phase: Each stream contains a number of samples defined there.
Number of Frames: Simulations are launched for a limited number of Frames.
Base of the Filename:
PhaseIn
by default, which generates the following names:Single Stream, Single Phase:
PhaseIn_0.txt
Single Stream, Polyphase:
PhaseIn_0.txt
,PhaseIn_1.txt
, …Dual streams, Polyphase:
PhaseIn_0_0.txt
,PhaseIn_0_0.txt
,PhaseIn_1_0.txt,
PhaseIn_1_0.txt``, …
Another possibility is to type GenerateStreams
with the same parameters. If you type GenerateStreams
without parameters, a usage text is displayed:
>>GenerateStreams
Stream Content Generation
================================================================================================
Usage:
GenerateStreams DataType PLIO_Width NPhases NStreams NSamples NFrames SequenceType Basename
Datatype: cint16, int16, int32
PLIO_Width: 32, 64 or 128
NPhases: any integer >= 1
NStreams: 1 or 2
NSamples: integer, multiple of 8
NFrames: Any integer >= 1
SequenceType: Dirac, Linear, SinCos, Random
Basename: file name that will prepend phase and stream index
================================================================================================