Custom Analyzers Reference & Examples

This section will provide more details of what you might need to write your own analyze() function. In the How to Import, Read, and Write a Data Group with VsHdf5 section below, several VsHdf5 class functions are listed which you will be able to use to import, read, and write data if you inherit from the VpAnalyzer class (Write a Class that Inherits from VpAnalyzer).

In Examples are examples of code you may find in the analyze() function.

How to Import, Read, and Write a Data Group with VsHdf5

The VpAnalyzer class uses VsHdf5 Python class functions for XSim file reading and writing. VsHdf5 ensures that metadata describing the type of datasets and other simulation information is added to simulation data so that it can be correctly visualized in Composer. XSim dumps simulation data in Hdf5 format with the extension “.h5”. It is good practice to make analyzers create files with the extension “.vsh5” in order to differentiate datasets that are simulation results versus those that are analysis results. This will protect against accidentally deleting simulation results that are not recoverable unless the simulation is re-run.

There are a number of convenient functions included in VpAnalyzer that handle VizSchema file reading and writing, and these should be used in your analyze() function. There should not be any need to import VsHdf5 directly into your analysis module, nor should you need to import PyTables. There are three types of functions. First, declaring Objects returns an empty object, (either a group or dataset) with the specified name. This is useful if you are creating a new dataset or group that will be written out to a file.

def History(name=None)
def Field(name=None)
def Limits(name=None)
def TimeGroup(name=None)
def Mesh(name=None, kind='uniformCartesian')
def RunInfo(name=None)
def Group(name=None)
def Dataset(name=None)

Second are functions for reading VsHdf5 objects from an existing file. This is useful for loading in XSim dump data.

def getHistory(fileName=None, historyName=None, name=None, location='/')
def getField(fileName=None, fieldName=None, name=None, location='/')
def getLimits(fileName=None, name=None)
def getTimeGroup(fileName=None, name=None)
def getMesh(fileName=None, name=None)
def getRunInfo(fileName=None, name=None)
def getGroup(fileName=None, groupName=None, name=None)
def getDataset(fileName=None, datasetName=None, name=None)

Finally, there are functions for writing VsHdf5 objects into a new or existing file.

def writeHistory(fileName=None, historyName=None, data=None, meshName=None, overwrite=False, location='/')
def writeField(fileName=None, data=None, fieldName=None, meshName=None, limitsName=None,    timeGroupName=None, offset='nodal', dumpTime=0.0, overwrite=False)
def writeLimits(fileName=None, limitsName=None, overwrite=False)
def writeTimeGroup(fileName=None, timeGroupName=None, dumpTime=None, dumpStep=None, overwrite=False)
def writeMesh(fileName=None, mesh=None, overwrite=False)
def writeRunInfo(fileName=None, runInfo=None, overwrite=False)
def writeGroup(fileName=None, groupName=None, overwrite=False)
def writeDataset(fileName=None, datasetName=None, overwrite=False)

Examples

Example 1: Create a 1D Data Set

Create a 1-D structured mesh and assign frequency values to that mesh’s dataset. Then define a History with values for each mesh point (frequency), and write that History to a file. Note that in this example, “self.overwrite” will be True if the --overwrite or -w flag was passed as a command line argument, or False otherwise. Similarly, --simulationName is a command line argument.

fileName = self.simulationName + '_SParameters.vsh5'
mesh = self.Mesh(name='SParameters', kind='structured')
mesh.assignDataset(Freq)
self.writeMesh(fileName=fileName, mesh=mesh, overwrite=self.overwrite)

pointDataHist = self.History(name='S11')
pointDataHist.assignDataset(S11)
pointDataHist.assignAttribute('vsCentering', 'nodal')
pointDataHist.writeHistory(fileName=fileName, meshName=mesh.name, overwrite=self.overwrite)

Example 2: Read Time Step and Dump Periodicity

Read a TimeGroup from a XSim dump file, and extract the simulation time and dump step from that group. Note that all of the get functions return both the object, and the attributes of that object.

timeGroup, attrs = self.getTimeGroup(fileName=speciesFileName)
time = timeGroup.getDumpTime()
step = timeGroup.getDumpStep()

Example 3: Extract Data from a History

Read a History from a XSim History dump file, and extract the data array from that History. Note that “hist” is a VsHdf5 object, not the actual dataset. To extract the dataset from the object, one can use the usual slicing operators, as in the example below, or by direct reference to the dataset, that is an attribute of the object: “data = hist.dataset”. data is a possibly multi-dimensional numpy array.

fileName = self.simulationName + '_History.h5'
hist, attrs = self.getHistory(fileName=fileName, historyName=self.historyName)
data = hist[:]