def convertFromDymolaMatFile(matFilename, mtsfFilename=None):
''' Converts a Dymola result file (in mat-format) into the MTSF format.
Returns the filename of the new result file
'''
# Define file name of result file
if mtsfFilename is None:
if len(matFilename) >= 4:
if matFilename[-4:] == '.mat':
resultFileName = matFilename[:-4] + '.mtsf'
else:
resultFileName = matFilename + '.mtsf'
else:
resultFileName = matFilename + '.mtsf'
else:
if '.mtsf' in mtsfFilename:
if mtsfFilename[-5:] == '.mtsf':
resultFileName = mtsfFilename
else:
resultFileName = mtsfFilename + '.mtsf'
else:
resultFileName = mtsfFilename + '.mtsf'
from ..DymolaMat import DymolaMat
# Load mat-file
res = DymolaMat.Results(matFilename)
# Define basic structure of result file
variable = collections.OrderedDict()
# Search for aliases
sortedVariables = [(i, res._dataInfo[i, 0], abs(res._dataInfo[i, 1]))
for i in xrange(len(res._name))]
sortedVariables.sort(key=itemgetter(2))
sortedVariables.sort(key=itemgetter(1))
aliasName = [None for i in xrange(len(res._name))]
for i, var in enumerate(sortedVariables):
index = var[0]
alias = None
j = i
while j > 0:
if sortedVariables[j - 1][1] != var[1] or sortedVariables[
j - 1][2] != var[2]:
break
else:
j -= 1
if j < i:
alias = res._name[sortedVariables[j][0]]
aliasName[index] = alias
dataIndexFixed = []
dataIndexContinuous = []
categoryIndex = pyMtsf.StandardCategoryNames.index(
pyMtsf.CategoryMapping['Real'])
for index, variableName in enumerate(res._name):
aliasNegated = False
if res._dataInfo[index, 0] == 1:
variability = 'fixed'
seriesIndex = 0 # Fixed
else:
variability = 'continuous'
seriesIndex = 1 # Continuous
if res._dataInfo[index, 1] < 0:
aliasNegated = True
if aliasName[index] is None:
if variability == 'fixed':
dataIndexFixed.append(abs(res._dataInfo[index, 1]) - 1)
else:
dataIndexContinuous.append(abs(res._dataInfo[index, 1]) - 1)
variable[variableName] = pyMtsf.ScalarModelVariable(
res._description[index],
'option',
0, # may be set later
variability,
seriesIndex,
categoryIndex,
aliasName[index],
aliasNegated)
modelVariables = pyMtsf.ModelVariables(variable,
MtsfFmi.StandardSeriesForFmi,
pyMtsf.StandardCategoryNames)
timeData = res.data("Time")
modelVariables.allSeries[1].initialRows = len(timeData) # Continuous
simpleTypes = []
units = []
enumerations = []
simpleTypes.append(
pyMtsf.SimpleType('Real without unit', pyMtsf.DataType["Real"], '',
False, -1, '')) # No unit
unitList = [(index, unit) for index, unit in enumerate(res._unit)]
unitList.sort(key=itemgetter(1))
preUnit = ''
for x in unitList:
index = x[0]
unit = x[1]
if unit != '':
if preUnit != unit:
units.append(pyMtsf.Unit(unit, 1.0, 0.0, 0))
simpleTypes.append(
pyMtsf.SimpleType('Real, Unit = ' + unit,
pyMtsf.DataType["Real"], '', False,
len(units) - 1, ''))
preUnit = unit
modelVariables.variable[
res._name[index]].simpleTypeRow = len(simpleTypes) - 1
else:
modelVariables.variable[
res._name[index]].simpleTypeRow = len(simpleTypes) - 1
experimentSetup = pyMtsf.ExperimentSetup(
startTime=timeData[0],
stopTime=timeData[-1],
algorithm="",
relativeTolerance='',
author="",
description="",
generationDateAndTime=time.strftime("%a, %d %b %Y %H:%M:%S",
time.gmtime()),
generationTool="Python",
machine=os.getenv('COMPUTERNAME'),
cpuTime="")
modelDescription = pyMtsf.ModelDescription(resultFileName[:-5], '', '', '',
'', '', 'structured')
# Create result object
mtsf = pyMtsf.MTSF(resultFileName, modelDescription, modelVariables,
experimentSetup, simpleTypes, units, enumerations)
# Write numeric data
fixedValues = numpy.ndarray((len(dataIndexFixed, )))
for i, index in enumerate(dataIndexFixed):
fixedValues[i] = res._data[0][0, index]
continuousValues = numpy.ndarray((len(timeData), len(dataIndexContinuous)))
for i, index in enumerate(dataIndexContinuous):
continuousValues[:, i] = res._data[1][:, index]
mtsf.results.series['Fixed'].category[
pyMtsf.CategoryMapping['Real']].writeData(fixedValues)
mtsf.results.series['Continuous'].category[
pyMtsf.CategoryMapping['Real']].writeData(continuousValues)
# Close file
mtsf.close()
return resultFileName
import time
import numpy
nPoints = 60
BlockSize = 100
# Prepare information from FMU
name_fmu_file = u'Examples/fullRobot'
(modelDescription, modelVariables, simpleTypes, units, enumerations) = convertFromFmi(name_fmu_file)
modelVariables.allSeries[1].initialRows = nPoints * BlockSize # Continuous
# Phase 1 of result file generation
resultFileName = name_fmu_file + unicode(nPoints) + u'.mtsf'
experimentSetup = pyMtsf.ExperimentSetup(startTime=0.0, stopTime=4.78, algorithm="Dassl",
relativeTolerance=1e-7, author="", description="Test experiment",
generationDateAndTime=time.strftime("%a, %d %b %Y %H:%M:%S", time.gmtime()),
generationTool="Python", machine=os.getenv('COMPUTERNAME'),
cpuTime="")
startTime = time.clock()
# Create result object
mtsf = pyMtsf.MTSF(resultFileName, modelDescription, modelVariables, experimentSetup, simpleTypes, units, enumerations)
# Some aliases
realParameter = mtsf.results.series['Fixed'].category[pyMtsf.CategoryMapping['Real']]
# integerParameter = mtsf.results.series['Fixed'].category[CategoryMapping['Integer']]
booleanParameter = mtsf.results.series['Fixed'].category[pyMtsf.CategoryMapping['Boolean']]
realContinuous = mtsf.results.series['Continuous'].category[pyMtsf.CategoryMapping['Real']]
realDiscrete = mtsf.results.series['Discrete'].category[pyMtsf.CategoryMapping['Real']]
# integerDiscrete = mtsf.results.series['Discrete'].category[CategoryMapping['Integer']]
booleanDiscrete = mtsf.results.series['Discrete'].category[pyMtsf.CategoryMapping['Boolean']]