PyTrilinos.Isorropia.NestedEpetra.Visualizer Class Reference

List of all members.

Public Member Functions
def	__init__
def	processRow
def	generateImage
def	show
def	save
def	getPILImage
def	simpleGenerate
def	setBgColor
def	setImageHeight
def	setImageWidth
def	setViewingWindow
def	setVerbosity
def	fillColorArray
def	centralizeMatrix
def	centralizePartitioner
def	centralizePartitioner2D
Public Attributes
	comm
	numProcs
	verbosity
	loadThreshold
	matrixHeight
	matrixWidth
	pointBR
	regionHeight
	regionWidth
	imageHeight
	fillColors
	background
	imageWidth
	pointTL
Static Public Attributes
	data = None
	dataType = None
int	numProcs = 0
int	verbosity = 0
	iAmRoot = False
int	loadThreshold = 0
	img = None
int	background = 0x000000
int	matrixHeight = 1
int	matrixWidth = 1
int	imageHeight = 1
int	imageWidth = 1
tuple	pointTL = (0,0)
tuple	pointBR = (-1,-1)
int	regionHeight = 1
int	regionWidth = 1

---

Detailed Description

The Visualizer class provides a means of simply visualizing partitions
formed by Isorropia.Epetra.Partitioner or Partitioner2D. It can also draw
Epetra.CrsMatrix objects directly.

The Visualizer relies on the Python Imaging Library and ImageMagick.

To generate images of multiple sizes given the same partitioner, you'll
need to call generateImage() several times, calling getPILImage() after
each call and saving the result. You can then save or show those images as
necessary.

---

Constructor & Destructor Documentation

def PyTrilinos.Isorropia.NestedEpetra.Visualizer.__init__	(		self,
			data,
			numParts = -1,
			loadThreshold = 0,
			verbosity = 0
	)

Visualizer constructor.
Inputs: data      - CrsMatrix or Partitioner
numParts  - The number of parts to distribute into.
            Defaults to the number of processes in MPI.
loadThreshold - The threshold ratio that must be exceeded
            for the visualizer to display something in that
            pixel. Zero is a safe choice.
verbosity - Print debugging output. High number increases
            quantity of output.

       The initialization first reassigns the values in the matrix to represent
       the processor they reside on (or have been partitioned to), then
       redistributes the matrix onto the root processor for use in show().

---

Member Function Documentation

def PyTrilinos.Isorropia.NestedEpetra.Visualizer.centralizeMatrix

(

self

)

Performs two operations:
    1. Replaces all non-zero values in the matrix with a value
       representing that non-zero's global part ID
    2. Repartitions such that all elements in the matrix are owned by
       the main process.

Reset the values in each row of the matrix to this processor's ID 

def PyTrilinos.Isorropia.NestedEpetra.Visualizer.centralizePartitioner

(

self

)

Performs three operations:
    1. Replaces the partitioner with a new CrsMatrix, maintaining
       non-zero distribution (with the same RowMap).
    1. Replaces all non-zero values in the matrix with a value
       representing that non-zero's global part ID
    2. Repartitions such that all elements in the matrix are owned by
       the main process.

def PyTrilinos.Isorropia.NestedEpetra.Visualizer.centralizePartitioner2D

(

self

)

Exactly the same as centralizePartitioner(), but accepting a Partitioner2D.

Creates a new CrsMatrix to represent the Partitioner2D's matrix
structure. This new CrsMatrix has the same distribution as the
Partitioner2D's matrix, but the values of the nonzeroes represent their
'destination' processors in the Partitioner's new Map

def PyTrilinos.Isorropia.NestedEpetra.Visualizer.fillColorArray	(		self,
			numParts
	)

Given the number of partitions, returns an array of at least that many
distinct color values

def PyTrilinos.Isorropia.NestedEpetra.Visualizer.generateImage

(

self

)

Opens an image window representing the matrix stored in this object's
data field. This opened window allows saving the image.

The colors chosen to represent each processor are evenly distributed
along the hue and presently can not be customized.

def PyTrilinos.Isorropia.NestedEpetra.Visualizer.getPILImage

(

self

)

Returns the PIL Image object generated by generateImage(). Returns None
if no image has yet been generated.

def PyTrilinos.Isorropia.NestedEpetra.Visualizer.processRow	(		self,
			row,
			draw
	)

processRow() is a private helper to draw an individual row of pixels of
the image when simpleGenerate() can't be used.

def PyTrilinos.Isorropia.NestedEpetra.Visualizer.save	(		self,
			filename
	)

Saves the pre-generated image to the given file. The format is derived
from the filename. Must be preceded by a call to generateImage().

def PyTrilinos.Isorropia.NestedEpetra.Visualizer.show

(

self

)

Shows the pre-generated image. Must be preceded by a call to generateImage().

def PyTrilinos.Isorropia.NestedEpetra.Visualizer.simpleGenerate	(		self,
			draw
	)

Draws the matrix to the provided image object, provided that the width
and height of the image exceed the number of columns and rows of the matrix.

Used by generateImage().

---

The documentation for this class was generated from the following file:

• NestedEpetra.py