def __retrieveDataHistory(self, attributes):
"""
Function to retrieve a History from the HDF5 database
@ In, attributes, dict, options (metadata muste be appended to the root group)
@ Out, tupleVar, tuple, tuple in which the first position is a dictionary of numpy arays (input variable)
and the second is a dictionary of the numpy arrays (output variables).
"""
# Check the outParam variables and the outputPivotVal filters
inParam, outParam, inputRow = attributes['inParam'], attributes[
'outParam'], copy.deepcopy(attributes.get('inputRow', None))
inputPivotVal, outputPivotVal = attributes.get(
'inputPivotValue', None), attributes.get('outputPivotValue', None)
pivotParameter = attributes.get('pivotParameter', None)
if 'all' in outParam: allOutParam = True
else: allOutParam = False
if outputPivotVal != None:
if 'all' in outputPivotVal: outputPivotValAll = True
else:
outputPivotValAll, outputPivotVal = False, [
float(x) for x in outputPivotVal.split()
]
else:
outputPivotValAll = True
if inputRow == None and inputPivotVal == None: inputRow = 0
if inputRow == None and inputPivotVal == None: inputRow = 0
if inputRow != None:
inputRow = int(inputRow)
if inputRow > 0: inputRow -= 1
inDict = {}
outDict = {}
metaDict = {}
# Call the function to retrieve a single history and
# load the data into the tuple
histVar = self.returnHistory(attributes)
if pivotParameter != None:
pivotIndex = histVar[1]['outputSpaceHeaders'].index(
pivotParameter
) if pivotParameter in histVar[1]['outputSpaceHeaders'] else None
if pivotIndex == None:
self.raiseAnError(
IOError, 'pivotParameter ' + pivotParameter +
' has not been found in Database ' + str(self.name) + '!')
else:
pivotIndex = histVar[1]['outputSpaceHeaders'].index(
"time") if "time" in histVar[1]['outputSpaceHeaders'] else None
# if None...default is 0
if pivotIndex == None: pivotIndex = 0
if 'metadata' in histVar[1].keys():
metaDict[0] = histVar[1]['metadata']
else:
metaDict[0] = None
# fill input param dictionary
for key in inParam:
if 'inputSpaceHeaders' in histVar[1]:
inInKey = utils.keyIn(histVar[1]['inputSpaceHeaders'], key)
inOutKey = utils.keyIn(histVar[1]['outputSpaceHeaders'], key)
if inInKey != None:
ix = histVar[1]['inputSpaceHeaders'].index(inInKey)
inDict[key] = np.atleast_1d(
np.array(histVar[1]['inputSpaceValues'][ix]))
elif inOutKey != None and inInKey == None:
ix = histVar[1]['outputSpaceHeaders'].index(inOutKey)
if inputPivotVal != None:
if float(inputPivotVal) > np.max(
histVar[0][:, pivotIndex]) or float(
inputPivotVal) < np.min(
histVar[0][:, pivotIndex]):
self.raiseAnError(
IOError,
'inputPivotVal is out of the min and max for input '
+ key + ' in Database ' + str(self.name) + '!')
inDict[key] = np.atleast_1d(
np.array(
interp1d(histVar[0][:, pivotIndex],
histVar[0][:, ix],
kind='linear')(float(inputPivotVal))))
else:
inDict[key] = np.atleast_1d(
np.array(histVar[0][inputRow, ix]))
else:
self.raiseAnError(
RuntimeError,
'the parameter ' + key + ' has not been found in ' +
str(histVar[1]['inputSpaceHeaders']) + ' or ' +
str(histVar[1]['outputSpaceHeaders']))
else:
inOutKey = utils.keyIn(histVar[1]['outputSpaceHeaders'], key)
if inOutKey is not None:
ix = histVar[1]['outputSpaceHeaders'].index(inOutKey)
if inputPivotVal != None:
if float(inputPivotVal) > np.max(
histVar[0][:, pivotIndex]) or float(
inputPivotVal) < np.min(
histVar[0][:, pivotIndex]):
self.raiseAnError(
IOError,
'inputPivotVal is out of the min and max for input '
+ key + ' in Database ' + str(self.name) + '!')
inDict[key] = np.atleast_1d(
np.array(
interp1d(histVar[0][:, pivotIndex],
histVar[0][:, ix],
kind='linear')(float(inputPivotVal))))
else:
inDict[key] = np.atleast_1d(
np.array(histVar[0][inputRow, ix]))
else:
self.raiseAnError(
RuntimeError,
'the parameter ' + key + ' has not been found in ' +
str(histVar[1]['outputSpaceHeaders']))
# all case => The history is completed (from startTime to end_time)
if outputPivotValAll:
if allOutParam:
for key in histVar[1]['outputSpaceHeaders']:
outDict[key] = histVar[0][:,
histVar[1]['outputSpaceHeaders'].
index(key)]
else:
for key in outParam:
inKey = utils.keyIn(histVar[1]['outputSpaceHeaders'], key)
if inKey:
outDict[key] = histVar[
0][:,
histVar[1]['outputSpaceHeaders'].index(inKey)]
else:
self.raiseAnError(
RuntimeError, 'the parameter ' + key +
' has not been found in ' +
str(histVar[1]['outputSpaceHeaders']))
else:
if allOutParam:
for key in histVar[1]['outputSpaceHeaders']:
outDict[key] = np.atleast_1d(
np.array(
interp1d(
histVar[0][:, pivotIndex],
histVar[0]
[:,
histVar[1]['outputSpaceHeaders'].index(key)],
kind='linear')(outputPivotVal)))
else:
for key in outParam:
if key in histVar[1]['outputSpaceHeaders']:
outDict[key] = np.atleast_1d(
np.array(
interp1d(
histVar[0][:, pivotIndex],
histVar[0][:,
histVar[1]['outputSpaceHeaders']
.index(key)],
kind='linear')(outputPivotVal)))
else:
self.raiseAnError(
IOError,
"the parameter " + key + " has not been found")
# Return tuple of dictionaries containing the HistorySet
return (copy.copy(inDict), copy.copy(outDict), copy.copy(metaDict))
def __retrieveDataPointSet(self, attributes):
"""
Function to retrieve a PointSet from the HDF5 database
@ In, attributes, dict, options (metadata must be appended to the root group)
@ Out, tupleVar, tuple, tuple in which the first position is a dictionary of numpy arays (input variable)
and the second is a dictionary of the numpy arrays (output variables).
"""
# Check the outParam variables and the outputPivotVal filters
inParam, outParam, inputRow, outputRow = attributes[
'inParam'], attributes['outParam'], copy.deepcopy(
attributes.get('inputRow', None)), copy.deepcopy(
attributes.get('outputRow', None))
inputPivotVal, outputPivotVal, operator = attributes.get(
'inputPivotValue',
None), attributes.get('outputPivotValue',
None), attributes.get('operator', None)
pivotParameter = attributes.get('pivotParameter', None)
if outParam == 'all': allOutParam = True
else: allOutParam = False
if outputPivotVal != None:
if 'end' in outputPivotVal: outputPivotValEnd = True
else:
outputPivotValEnd, outputPivotVal = False, float(
outputPivotVal)
else:
if operator is None: outputPivotValEnd = True
else: outputPivotValEnd = False
if inputRow == None and inputPivotVal == None: inputRow = 0
if inputRow == None and inputPivotVal == None: inputRow = 0
if inputRow != None:
inputRow = int(inputRow)
if inputRow > 0: inputRow -= 1
if outputRow != None:
outputRow = int(outputRow)
if outputRow > 0: outputRow -= 1
inDict = {}
outDict = {}
metaDict = {}
histList = attributes['HistorySet']
# Retrieve all the associated HistorySet and process them
for i in range(len(histList)):
# Load the data into the numpy array
attributes['history'] = histList[i]
histVar = self.returnHistory(attributes)
#look for pivotParameter
if pivotParameter != None:
pivotIndex = histVar[1]['outputSpaceHeaders'].index(
pivotParameter) if pivotParameter in histVar[1][
'outputSpaceHeaders'] else None
if pivotIndex == None:
self.raiseAnError(
IOError, 'pivotParameter ' + pivotParameter +
' has not been found in history ' +
str(attributes['history']) + '!')
else:
pivotIndex = histVar[1]['outputSpaceHeaders'].index(
"time"
) if "time" in histVar[1]['outputSpaceHeaders'] else None
# if None...default is 0
if pivotIndex == None: pivotIndex = 0
if inputRow > histVar[0][:, 0].size - 1 and inputRow != -1:
self.raiseAnError(
IOError,
'inputRow is greater than number of actual rows in history '
+ str(attributes['history']) + '!')
# check metadata
if 'metadata' in histVar[1].keys():
metaDict[i] = histVar[1]['metadata']
else:
metaDict[i] = None
for key in inParam:
if 'inputSpaceHeaders' in histVar[1]:
inInKey = utils.keyIn(histVar[1]['inputSpaceHeaders'], key)
inOutKey = utils.keyIn(histVar[1]['outputSpaceHeaders'],
key)
if inInKey != None:
ix = histVar[1]['inputSpaceHeaders'].index(inInKey)
if i == 0: inDict[key] = np.zeros(len(histList))
inDict[key][i] = np.atleast_1d(
histVar[1]['inputSpaceValues'][ix])[0]
elif inOutKey != None and inInKey == None:
ix = histVar[1]['outputSpaceHeaders'].index(inOutKey)
if i == 0: inDict[key] = np.zeros(len(histList))
if inputPivotVal != None:
if float(inputPivotVal) > np.max(
histVar[0][:, pivotIndex]) or float(
inputPivotVal) < np.min(
histVar[0][:, pivotIndex]):
self.raiseAnError(
IOError,
'inputPivotVal is out of the min and max for input '
+ key + ' in Database ' + str(self.name) +
'!')
inDict[key][i] = interp1d(
histVar[0][:, pivotIndex],
histVar[0][:, ix],
kind='linear')(float(inputPivotVal))
else:
inDict[key][i] = histVar[0][inputRow, ix]
else:
self.raiseAnError(
IOError,
'the parameter ' + key + ' has not been found')
else:
inKey = utils.keyIn(histVar[1]['outputSpaceHeaders'], key)
if inKey is not None:
ix = histVar[1]['outputSpaceHeaders'].index(inKey)
if i == 0: inDict[key] = np.zeros(len(histList))
if inputPivotVal != None:
if float(inputPivotVal) > np.max(
histVar[0][:, pivotIndex]) or float(
inputPivotVal) < np.min(
histVar[0][:, pivotIndex]):
self.raiseAnError(
IOError,
'inputPivotVal is out of the min and max for input '
+ key + ' in Database ' + str(self.name) +
'!')
inDict[key][i] = interp1d(
histVar[0][:, pivotIndex],
histVar[0][:, ix],
kind='linear')(float(inputPivotVal))
else:
inDict[key][i] = histVar[0][inputRow, ix]
else:
self.raiseAnError(
IOError, 'the parameter ' + key +
' has not been found in ' + str(histVar[1]))
# outputPivotVal end case => PointSet is at the final status
if outputPivotValEnd:
if allOutParam:
for key in histVar[1]['outputSpaceHeaders']:
if i == 0: outDict[key] = np.zeros(len(histList))
outDict[key][i] = histVar[0][
-1, histVar[1]['outputSpaceHeaders'].index(key)]
else:
for key in outParam:
if key in histVar[1]['outputSpaceHeaders'] or \
utils.toBytes(key) in histVar[1]['outputSpaceHeaders']:
if i == 0: outDict[key] = np.zeros(len(histList))
if key in histVar[1]['outputSpaceHeaders']:
outDict[key][i] = histVar[0][
-1, histVar[1]['outputSpaceHeaders'].
index(key)]
else:
outDict[key][i] = histVar[0][
-1, histVar[1]['outputSpaceHeaders'].
index(utils.toBytes(key))]
else:
self.raiseAnError(
IOError, "the parameter " + str(key) +
" has not been found")
elif outputRow != None:
if outputRow > histVar[0][:, 0].size - 1 and outputRow != -1:
self.raiseAnError(
IOError,
'outputRow is greater than number of actual rows in Database '
+ str(self.name) + '!')
if allOutParam:
for key in histVar[1]['outputSpaceHeaders']:
if i == 0: outDict[key] = np.zeros(len(histList))
outDict[key][i] = histVar[0][
outputRow,
histVar[1]['outputSpaceHeaders'].index(key)]
else:
for key in outParam:
if key in histVar[1]['outputSpaceHeaders'] or \
utils.toBytes(key) in histVar[1]['outputSpaceHeaders']:
if i == 0: outDict[key] = np.zeros(len(histList))
if key in histVar[1]['outputSpaceHeaders']:
outDict[key][i] = histVar[0][
outputRow,
histVar[1]['outputSpaceHeaders'].index(key
)]
else:
outDict[key][i] = histVar[0][
outputRow,
histVar[1]['outputSpaceHeaders'].
index(utils.toBytes(key))]
else:
self.raiseAnError(
IOError, "the parameter " + str(key) +
" has not been found")
elif operator != None:
if operator not in ['max', 'min', 'average']:
self.raiseAnError(
IOError,
'operator unknown. Available are min,max,average')
if histVar[1]['outputSpaceHeaders']:
for key in histVar[1]['outputSpaceHeaders']:
if i == 0: outDict[key] = np.zeros(len(histList))
if operator == 'max':
outDict[key][i] = np.max(
histVar[0]
[:,
histVar[1]['outputSpaceHeaders'].index(key)])
if operator == 'min':
outDict[key][i] = np.min(
histVar[0]
[:,
histVar[1]['outputSpaceHeaders'].index(key)])
if operator == 'average':
outDict[key][i] = np.average(
histVar[0]
[:,
histVar[1]['outputSpaceHeaders'].index(key)])
else:
for key in outParam:
if key in histVar[1]['outputSpaceHeaders'] or \
utils.toBytes(key) in histVar[1]['outputSpaceHeaders']:
if i == 0: outDict[key] = np.zeros(len(histList))
if key in histVar[1]['outputSpaceHeaders']:
if operator == 'max':
outDict[key][i] = np.max(
histVar[0]
[:, histVar[1]['outputSpaceHeaders'].
index(key)])
if operator == 'min':
outDict[key][i] = np.min(
histVar[0]
[:, histVar[1]['outputSpaceHeaders'].
index(key)])
if operator == 'average':
outDict[key][i] = np.average(
histVar[0]
[:, histVar[1]['outputSpaceHeaders'].
index(key)])
else:
if operator == 'max':
outDict[key][i] = np.max(
histVar[0]
[:, histVar[1]['outputSpaceHeaders'].
index(utils.toBytes(key))])
if operator == 'min':
outDict[key][i] = np.min(
histVar[0]
[:, histVar[1]['outputSpaceHeaders'].
index(utils.toBytes(key))])
if operator == 'average':
outDict[key][i] = np.average(
histVar[0]
[:, histVar[1]['outputSpaceHeaders'].
index(utils.toBytes(key))])
else:
self.raiseAnError(
IOError, "the parameter " + str(key) +
" has not been found")
else:
# Arbitrary point in outputPivotVal case... If the requested outputPivotVal point Set does not match any of the stored ones and
# start_outputPivotVal <= requested_outputPivotVal_point <= end_outputPivotVal, compute an interpolated value
if allOutParam:
for key in histVar[1]['outputSpaceHeaders']:
if i == 0: outDict[key] = np.zeros(len(histList))
outDict[key][i] = interp1d(
histVar[0][:, pivotIndex],
histVar[0]
[:, histVar[1]['outputSpaceHeaders'].index(key)],
kind='linear')(outputPivotVal)
else:
for key in outParam:
if i == 0: outDict[key] = np.zeros(len(histList))
if key in histVar[1]['outputSpaceHeaders'] or \
utils.toBytes(key) in histVar[1]['outputSpaceHeaders']:
if key in histVar[1]['outputSpaceHeaders']:
outDict[key][i] = interp1d(
histVar[0][:, pivotIndex],
histVar[0][:,
histVar[1]['outputSpaceHeaders']
.index(key)],
kind='linear')(outputPivotVal)
else:
outDict[key][i] = interp1d(
histVar[0][:, pivotIndex],
histVar[0][:,
histVar[1]['outputSpaceHeaders']
.index(utils.toBytes(key))],
kind='linear')(outputPivotVal)
else:
self.raiseAnError(
IOError,
"the parameter " + key + " has not been found")
del histVar
# return tuple of PointSet
return (copy.copy(inDict), copy.copy(outDict), copy.copy(metaDict))