def errvar_test():
import os
from pyemu import ErrVar
w_dir = os.path.join("..", "verification", "henry")
forecasts = ["pd_ten", "c_obs10_2"]
ev = ErrVar(jco=os.path.join(w_dir, "pest.jcb"), forecasts=forecasts)
print(ev.prior_forecast)
print(ev.get_errvar_dataframe())
def errvar_test():
import os
from pyemu import ErrVar
w_dir = os.path.join("..","..","verification","henry")
forecasts = ["pd_ten","c_obs10_2"]
ev = ErrVar(jco=os.path.join(w_dir,"pest.jcb"),forecasts=forecasts)
print(ev.prior_forecast)
print(ev.get_errvar_dataframe())
def __init__(self, jco, par_info_file=None):
"""Computes parameter identifiability for a PEST jco file,
using the ErrVar class in pyemu (https://github.com/jtwhite79/pyemu)
"""
self._Pest = Pest(jco, par_info_file=par_info_file)
self.parinfo = self._Pest.parinfo
self.la = ErrVar(jco)
self.parinfo = None
if par_info_file is not None:
self.parinfo = pd.read_csv(par_info_file, index_col='Name')
self.ident_df = None
def errvar_test_nonpest():
import numpy as np
from pyemu import ErrVar, Matrix, Cov
#non-pest
pnames = ["p1","p2","p3"]
onames = ["o1","o2","o3","o4"]
npar = len(pnames)
nobs = len(onames)
j_arr = np.random.random((nobs,npar))
jco = Matrix(x=j_arr,row_names=onames,col_names=pnames)
parcov = Cov(x=np.eye(npar),names=pnames)
obscov = Cov(x=np.eye(nobs),names=onames)
forecasts = "o2"
omitted = "p3"
e = ErrVar(jco=jco,parcov=parcov,obscov=obscov,forecasts=forecasts,
omitted_parameters=omitted)
svs = [0,1,2,3,4,5]
print(e.get_errvar_dataframe(svs))
def errvar_test_nonpest():
import numpy as np
from pyemu import ErrVar, Matrix, Cov
#non-pest
pnames = ["p1","p2","p3"]
onames = ["o1","o2","o3","o4"]
npar = len(pnames)
nobs = len(onames)
j_arr = np.random.random((nobs,npar))
jco = Matrix(x=j_arr,row_names=onames,col_names=pnames)
parcov = Cov(x=np.eye(npar),names=pnames)
obscov = Cov(x=np.eye(nobs),names=onames)
forecasts = "o2"
omitted = "p3"
e = ErrVar(jco=jco,parcov=parcov,obscov=obscov,forecasts=forecasts,
omitted_parameters=omitted)
svs = [0,1,2,3,4,5]
print(e.get_errvar_dataframe(svs))
def __init__(self, jco, par_info_file=None):
"""Computes parameter identifiability for a PEST jco file,
using the ErrVar class in pyemu (https://github.com/jtwhite79/pyemu)
"""
self._Pest = Pest(jco, par_info_file=par_info_file)
self.parinfo = self._Pest.parinfo
self.la = ErrVar(jco)
self.parinfo = None
if par_info_file is not None:
self.parinfo = pd.read_csv(par_info_file, index_col='Name')
self.ident_df = None
class IdentPar:
def __init__(self, jco, par_info_file=None):
"""Computes parameter identifiability for a PEST jco file,
using the ErrVar class in pyemu (https://github.com/jtwhite79/pyemu)
"""
self._Pest = Pest(jco, par_info_file=par_info_file)
self.parinfo = self._Pest.parinfo
self.la = ErrVar(jco)
self.parinfo = None
if par_info_file is not None:
self.parinfo = pd.read_csv(par_info_file, index_col='Name')
self.ident_df = None
def plot_singular_spectrum(self):
"""see http://nbviewer.ipython.org/github/jtwhite79/pyemu/blob/master/examples/error_variance_example.ipynb
"""
s = self.la.qhalfx.s
figure = plt.figure(figsize=(10, 5))
ax = plt.subplot(111)
ax.plot(s.x)
ax.set_title("singular spectrum")
ax.set_ylabel("power")
ax.set_xlabel("singular value")
ax.set_xlim(0, 300)
ax.set_ylim(0, 10)
plt.show()
def get_identifiability_dataframe(self, nsingular):
self.ident_df = self.la.get_identifiability_dataframe(nsingular)
if self.parinfo is not None:
self.ident_points = pd.DataFrame({
'ident_sum':
self.ident_df.sum(axis=1)
}).join(self.parinfo)
self.ident_points.ident_sum = [
i / 2.0 for i in self.ident_points.ident_sum
]
def plot_bar(self, nsingular=None, nbars=20):
"""Computes a stacked bar chart showing the most identifiable parameters
at a given number of singular values
Parameters:
-----------
nsingular:
number of singular values to include
nbars:
number of parameters (bars) to include in bar chart
"""
if nsingular is not None:
self.get_identifiability_dataframe(nsingular)
plot_obj = plots.IdentBar(self.ident_df,
nsingular=nsingular,
nbars=nbars)
plot_obj.generate()
plot_obj.draw()
return plot_obj.fig, plot_obj.ax
def plot_spatial(self,
nsingular=None,
groupinfo={},
group_col=None,
legend_kwds={},
**kwds):
if nsingular is not None:
self.get_identifiability_dataframe(nsingular)
"""
Make a spatial plot of total parameter identifiabilities
Parameters
----------
Notes
------
"""
# if no group column is argued, default to all spatial parameters in one group
df = self.ident_points
if group_col is None:
group_col = 'partype'
df.loc[np.isnan(df.X), group_col] = 'lumped'
df.loc[~np.isnan(df.X), group_col] = 'spatial'
plot_obj = plots.SpatialPlot(df,
'X',
'Y',
'ident_sum',
groupinfo,
group_col=group_col,
colorby='k',
legend_values=[0, 0.5, 1],
legend_kwds=legend_kwds,
**kwds)
plot_obj.generate()
plot_obj.draw()
return plot_obj
class IdentPar:
def __init__(self, jco, par_info_file=None):
"""Computes parameter identifiability for a PEST jco file,
using the ErrVar class in pyemu (https://github.com/jtwhite79/pyemu)
"""
self._Pest = Pest(jco, par_info_file=par_info_file)
self.parinfo = self._Pest.parinfo
self.la = ErrVar(jco)
self.parinfo = None
if par_info_file is not None:
self.parinfo = pd.read_csv(par_info_file, index_col='Name')
self.ident_df = None
def plot_singular_spectrum(self):
"""see http://nbviewer.ipython.org/github/jtwhite79/pyemu/blob/master/examples/error_variance_example.ipynb
"""
s = self.la.qhalfx.s
figure = plt.figure(figsize=(10, 5))
ax = plt.subplot(111)
ax.plot(s.x)
ax.set_title("singular spectrum")
ax.set_ylabel("power")
ax.set_xlabel("singular value")
ax.set_xlim(0,300)
ax.set_ylim(0,10)
plt.show()
def get_identifiability_dataframe(self, nsingular):
self.ident_df = self.la.get_identifiability_dataframe(nsingular)
if self.parinfo is not None:
self.ident_points = pd.DataFrame({'ident_sum': self.ident_df.sum(axis=1)}).join(self.parinfo)
self.ident_points.ident_sum = [i/2.0 for i in self.ident_points.ident_sum]
def plot_bar(self, nsingular=None, nbars=20):
"""Computes a stacked bar chart showing the most identifiable parameters
at a given number of singular values
Parameters:
-----------
nsingular:
number of singular values to include
nbars:
number of parameters (bars) to include in bar chart
"""
if nsingular is not None:
self.get_identifiability_dataframe(nsingular)
plot_obj = plots.IdentBar(self.ident_df, nsingular=nsingular, nbars=nbars)
plot_obj.generate()
plot_obj.draw()
return plot_obj.fig, plot_obj.ax
def plot_spatial(self, nsingular=None,
groupinfo={},
group_col=None,
legend_kwds={}, **kwds):
if nsingular is not None:
self.get_identifiability_dataframe(nsingular)
"""
Make a spatial plot of total parameter identifiabilities
Parameters
----------
Notes
------
"""
# if no group column is argued, default to all spatial parameters in one group
df = self.ident_points
if group_col is None:
group_col = 'partype'
df.loc[np.isnan(df.X), group_col] = 'lumped'
df.loc[~np.isnan(df.X), group_col] = 'spatial'
plot_obj = plots.SpatialPlot(df, 'X', 'Y', 'ident_sum', groupinfo, group_col=group_col,
colorby='k', legend_values=[0, 0.5, 1],
legend_kwds=legend_kwds, **kwds)
plot_obj.generate()
plot_obj.draw()
return plot_obj
