class ComputeExtrap(object):
"""Class to compute the optimized or manually specified extrapolation methods
Attributes
----------
threshold: float
Threshold as a percent for determining if a median is valid
subsection: list
Percent of discharge, does not account for transect direction
fit_method: str
Method used to determine fit. Automatic or manual
norm_data: NormData
Object of class NormData
sel_fit: SelectFit
Object of class SelectFit
q_sensitivity: ExtrapQSensitivity
Object of class ExtrapQSensitivity
messages: str
Variable for messages to UserWarning
"""
def __init__(self):
"""Initialize instance variables."""
self.threshold = None # Threshold as a percent for determining if a median is valid
self.subsection = None # Percent of discharge, does not account for transect direction
self.fit_method = None # Method used to determine fit. Automatic or manual
self.norm_data = [] # Object of class norm data
self.sel_fit = [] # Object of class SelectFit
self.q_sensitivity = None # Object of class ExtrapQSensitivity
self.messages = [] # Variable for messages to UserWarning
def populate_data(self, transects, compute_sensitivity=True):
"""Store data in instance variables.
Parameters
----------
transects: list
List of transects of TransectData
compute_sensitivity: bool
Determines is sensitivity should be computed.
"""
self.threshold = 20
self.subsection = [0, 100]
self.fit_method = 'Automatic'
self.process_profiles(transects=transects, data_type='q')
# Compute the sensitivity of the final discharge to changes in extrapolation methods
if compute_sensitivity:
self.q_sensitivity = ExtrapQSensitivity()
self.q_sensitivity.populate_data(transects=transects,
extrap_fits=self.sel_fit)
def process_profiles(self, transects, data_type):
"""Function that coordinates the fitting process.
Parameters
----------
transects: TransectData
Object of TransectData
data_type: str
Type of data processing (q or v)
"""
# Compute normalized data for each transect
self.norm_data = []
for transect in transects:
norm_data = NormData()
norm_data.populate_data(transect=transect,
data_type=data_type,
threshold=self.threshold,
data_extent=self.subsection)
self.norm_data.append(norm_data)
# Compute composite normalized data
comp_data = NormData()
comp_data.create_composite(transects=transects,
norm_data=self.norm_data,
threshold=self.threshold)
self.norm_data.append(comp_data)
sel_fit = None
# Compute the fit for the selected method
self.sel_fit = []
if self.fit_method == 'Manual':
for n in range(len(transects)):
sel_fit = SelectFit()
sel_fit.populate_data(normalized=self.norm_data[n],
fit_method=self.fit_method,
transect=transects[n])
self.sel_fit.append(sel_fit)
else:
for n in range(len(self.norm_data)):
sel_fit = SelectFit()
sel_fit.populate_data(self.norm_data[n], self.fit_method)
self.sel_fit.append(sel_fit)
if sel_fit.top_fit_r2 is not None:
# Evaluate if there is a potential that a 3-point top method may be appropriate
if (sel_fit.top_fit_r2 > 0.9 or sel_fit.top_r2 > 0.9) and np.abs(
sel_fit.top_max_diff) > 0.2:
self.messages.append(
'The measurement profile may warrant a 3-point fit at the top'
)
def update_q_sensitivity(self, transects):
self.q_sensitivity = ExtrapQSensitivity()
self.q_sensitivity.populate_data(transects, self.sel_fit)
def change_fit_method(self, transect, new_fit_method, n, kargs=None):
# TODO this function needs to be thought through only appears to be needed for setting individual transects to view in extrap window
"""Function to change the extrapolation methods associated with single transect"""
self.fit_method = new_fit_method
self.sel_fit = SelectFit()
self.sel_fit.populate_data(self.norm_data, new_fit_method, kargs)
# self.q_sensitivity = ExtrapQSensitivity()
# self.q_sensitivity.populate_data(trans_data, self.sel_fit)
def change_threshold(self, transects, data_type, threshold):
"""Function to change the threshold for accepting the increment median as valid. The threshold
is in percent of the median number of points in all increments"""
self.threshold = threshold
self.process_profiles(transects=transects, data_type=data_type)
self.q_sensitivity = ExtrapQSensitivity()
self.q_sensitivity.populate_data(transects=transects,
extrap_fits=self.sel_fit)
def change_extents(self, transects, data_type, extents):
"""Function allows the data to be subsection by specifying the percent cumulative discharge
for the start and end points. Currently this function does not consider transect direction"""
self.subsection = extents
self.process_profiles(transects=transects, data_type=data_type)
self.q_sensitivity = ExtrapQSensitivity()
self.q_sensitivity.populate_data(transects=transects,
extrap_fits=self.sel_fit)
def change_data_type(self, trans_data, data_type):
self.process_profiles(trans_data, data_type)
self.q_sensitivity = ExtrapQSensitivity(trans_data, self.selfit)
class ComputeExtrap(object):
"""Class to compute the optimized or manually specified extrapolation methods
Attributes
----------
threshold: float
Threshold as a percent for determining if a median is valid
subsection: list
Percent of discharge, does not account for transect direction
fit_method: str
Method used to determine fit. Automatic or manual
norm_data: NormData
Object of class NormData
sel_fit: SelectFit
Object of class SelectFit
q_sensitivity: ExtrapQSensitivity
Object of class ExtrapQSensitivity
messages: str
Variable for messages to UserWarning
"""
def __init__(self):
"""Initialize instance variables."""
self.threshold = None # Threshold as a percent for determining if a median is valid
self.subsection = None # Percent of discharge, does not account for transect direction
self.fit_method = None # Method used to determine fit. Automatic or manual
self.norm_data = [] # Object of class norm data
self.sel_fit = [] # Object of class SelectFit
self.q_sensitivity = None # Object of class ExtrapQSensitivity
self.messages = [] # Variable for messages to UserWarning
def populate_data(self, transects, compute_sensitivity=True):
"""Store data in instance variables.
Parameters
----------
transects: list
List of transects of TransectData
compute_sensitivity: bool
Determines is sensitivity should be computed.
"""
self.threshold = 20
self.subsection = [0, 100]
self.fit_method = 'Automatic'
self.process_profiles(transects=transects, data_type='q')
# Compute the sensitivity of the final discharge to changes in extrapolation methods
if compute_sensitivity:
self.q_sensitivity = ExtrapQSensitivity()
self.q_sensitivity.populate_data(transects=transects,
extrap_fits=self.sel_fit)
def populate_from_qrev_mat(self, meas_struct):
"""Populates the object using data from previously saved QRev Matlab file.
Parameters
----------
meas_struct: mat_struct
Matlab data structure obtained from sio.loadmat
"""
if hasattr(meas_struct, 'extrapFit'):
self.threshold = meas_struct.extrapFit.threshold
self.subsection = meas_struct.extrapFit.subsection
self.fit_method = meas_struct.extrapFit.fitMethod
self.norm_data = NormData.qrev_mat_in(meas_struct.extrapFit)
self.sel_fit = SelectFit.qrev_mat_in(meas_struct.extrapFit)
self.q_sensitivity = ExtrapQSensitivity()
self.q_sensitivity.populate_from_qrev_mat(meas_struct.extrapFit)
if type(meas_struct.extrapFit.messages) is str:
self.messages = [meas_struct.extrapFit.messages]
elif type(meas_struct.extrapFit.messages) is np.ndarray:
self.messages = meas_struct.extrapFit.messages.tolist()
def process_profiles(self, transects, data_type):
"""Function that coordinates the fitting process.
Parameters
----------
transects: TransectData
Object of TransectData
data_type: str
Type of data processing (q or v)
"""
# Compute normalized data for each transect
self.norm_data = []
for transect in transects:
norm_data = NormData()
norm_data.populate_data(transect=transect,
data_type=data_type,
threshold=self.threshold,
data_extent=self.subsection)
self.norm_data.append(norm_data)
# Compute composite normalized data
comp_data = NormData()
comp_data.create_composite(transects=transects,
norm_data=self.norm_data,
threshold=self.threshold)
self.norm_data.append(comp_data)
# Compute the fit for the selected method
if self.fit_method == 'Manual':
for n in range(len(transects)):
self.sel_fit[n].populate_data(
normalized=self.norm_data[n],
fit_method=self.fit_method,
top=transects[n].extrap.top_method,
bot=transects[n].extrap.bot_method,
exponent=transects[n].extrap.exponent)
else:
self.sel_fit = []
for n in range(len(self.norm_data)):
sel_fit = SelectFit()
sel_fit.populate_data(self.norm_data[n], self.fit_method)
self.sel_fit.append(sel_fit)
if self.sel_fit[-1].top_fit_r2 is not None:
# Evaluate if there is a potential that a 3-point top method may be appropriate
if (self.sel_fit[-1].top_fit_r2 > 0.9 or self.sel_fit[-1].top_r2 > 0.9) \
and np.abs(self.sel_fit[-1].top_max_diff) > 0.2:
self.messages.append(
'The measurement profile may warrant a 3-point fit at the top'
)
def update_q_sensitivity(self, transects):
"""Updates the discharge sensitivity values.
Parameters
----------
transects: list
List of TransectData objects
"""
self.q_sensitivity = ExtrapQSensitivity()
self.q_sensitivity.populate_data(transects, self.sel_fit)
def change_fit_method(self,
transects,
new_fit_method,
idx,
top=None,
bot=None,
exponent=None,
compute_qsens=True):
"""Function to change the extrapolation method.
Parameters
----------
transects: list
List of TransectData objects
new_fit_method: str
Identifies fit method automatic or manual
idx: int
Index to the specified transect or measurement in NormData
top: str
Specifies top fit
bot: str
Specifies bottom fit
exponent: float
Specifies exponent for power or no slip fits
compute_qsens: bool
Specifies if the discharge sensitivities should be recomputed
"""
self.fit_method = new_fit_method
self.sel_fit[idx].populate_data(self.norm_data[idx],
new_fit_method,
top=top,
bot=bot,
exponent=exponent)
if compute_qsens & idx == len(self.norm_data) - 1:
self.q_sensitivity = ExtrapQSensitivity()
self.q_sensitivity.populate_data(transects, self.sel_fit)
def change_threshold(self, transects, data_type, threshold):
"""Function to change the threshold for accepting the increment median as valid. The threshold
is in percent of the median number of points in all increments.
Parameters
----------
transects: list
List of TransectData objects
data_type: str
Specifies the data type (discharge or velocity)
threshold: float
Percent of data that must be in a median to include the median in the fit algorithm
"""
self.threshold = threshold
self.process_profiles(transects=transects, data_type=data_type)
self.q_sensitivity = ExtrapQSensitivity()
self.q_sensitivity.populate_data(transects=transects,
extrap_fits=self.sel_fit)
def change_extents(self, transects, data_type, extents):
"""Function allows the data to be subsection by specifying the percent cumulative discharge
for the start and end points. Currently this function does not consider transect direction.
Parameters
----------
transects: list
List of TransectData objects
data_type: str
Specifies the data type (discharge or velocity)
extents: list
List containing two values, the minimum and maximum discharge percentages to subsectioning
"""
self.subsection = extents
self.process_profiles(transects=transects, data_type=data_type)
self.q_sensitivity = ExtrapQSensitivity()
self.q_sensitivity.populate_data(transects=transects,
extrap_fits=self.sel_fit)
def change_data_type(self, transects, data_type):
"""Changes the data type to be processed in extrap.
Parameters
----------
transects: list
List of TransectData objects
data_type: str
Specifies the data type (discharge or velocity)
"""
self.process_profiles(transects=transects, data_type=data_type)
self.q_sensitivity = ExtrapQSensitivity()
self.q_sensitivity.populate_data(transects=transects,
extrap_fits=self.sel_fit)
def change_data_auto(self, transects):
"""Changes the data selection settings to automatic.
Parameters
----------
transects: list
List of TransectData objects
"""
self.threshold = 20
self.subsection = [0, 100]
self.process_profiles(transects=transects, data_type='q')
# Compute the sensitivity of the final discharge to changes in extrapolation methods
self.q_sensitivity = ExtrapQSensitivity()
self.q_sensitivity.populate_data(transects=transects,
extrap_fits=self.sel_fit)