# Diff of /trunk/downunder/py_src/datasources.py

revision 4019 by jfenwick, Thu Oct 11 08:12:55 2012 UTC revision 4108 by caltinay, Thu Dec 13 06:38:11 2012 UTC
24
25  __all__ = ['DataSource','UBCDataSource','ERSDataSource','SyntheticDataSource','SmoothAnomaly']  __all__ = ['simpleBackgroundMagneticField', 'DataSource','ErMapperData','SyntheticFeatureData','SmoothAnomaly']
26
27  import logging  import logging
28  import numpy as np  import numpy as np
29  from esys.escript import ReducedFunction, Scalar  from esys.escript import ReducedFunction
30  from esys.escript.linearPDEs import LinearSinglePDE  from esys.escript.linearPDEs import LinearSinglePDE
31  from esys.escript.util import *  from esys.escript.util import *
32  import esys.escript.unitsSI as U  import esys.escript.unitsSI as U
33  from esys.ripley import Brick, Rectangle, ripleycpp  from esys.ripley import Brick, Rectangle, ripleycpp
import sys

if sys.version_info.major>2:
xrange=range
34
35  try:  try:
36      from scipy.io.netcdf import netcdf_file      from scipy.io.netcdf import netcdf_file
37      __all__ += ['NetCDFDataSource']      __all__ += ['NetCdfData']
38  except:  except:
39      pass      pass
40
# Line 47  def LatLonToUTM(lon, lat, wkt_string=Non Line 43  def LatLonToUTM(lon, lat, wkt_string=Non
43      Converts one or more longitude,latitude pairs to the corresponding x,y      Converts one or more longitude,latitude pairs to the corresponding x,y
44      coordinates in the Universal Transverse Mercator projection.      coordinates in the Universal Transverse Mercator projection.
45
46      :note: If the ``pyproj`` module is not installed a warning if printed and      :note: If the ``pyproj`` module is not installed a warning is printed and
47             the input values are scaled by a constant and returned.             the input values are scaled by a constant and returned.
48      :note: If `wkt_string` is not given or invalid or the ``gdal`` module is      :note: If `wkt_string` is not given or invalid or the ``gdal`` module is
49             not available to convert the string, then the input values are             not available to convert the string, then the input values are
# Line 82  def LatLonToUTM(lon, lat, wkt_string=Non Line 78  def LatLonToUTM(lon, lat, wkt_string=Non
78      x,y=pyproj.transform(p_src, p_dest, lon, lat)      x,y=pyproj.transform(p_src, p_dest, lon, lat)
79      return x,y      return x,y
80
81    def simpleBackgroundMagneticField(latitude, longitude=0.):
82            theta = (90-latitude)/180.*np.pi
83            B_0=U.Mu_0  * U.Magnetic_Dipole_Moment_Earth / (4 * np.pi *  U.R_Earth**3)
84            B_theta= B_0 * sin(theta)
85            B_r= 2 * B_0 * cos(theta)
86            return B_r, B_theta, 0.
87
88  class DataSource(object):  class DataSource(object):
89      """      """
90      A class that provides survey data for the inversion process.      A class that provides survey data for the inversion process.
91      This is an abstract base class that implements common functionality.      This is an abstract base class that implements common functionality.
92      Methods to be overwritten by subclasses are marked as such.      Methods to be overwritten by subclasses are marked as such.
93        This class assumes 2D data which is mapped to a slice of a 3D domain.
94        For other setups override the methods as required.
95      """      """
96      # this is currently specific to gravity inversion and should be generalised
97        GRAVITY, MAGNETIC = list(range(2))
98
99      def __init__(self):      def __init__(self):
100          """          """
101          Constructor. Sets some defaults and initializes logger.          Constructor. Sets some defaults and initializes logger.
102          """          """
self._constrainBottom=False
self._constrainSides=True
self._domain=None
103          self.logger = logging.getLogger('inv.%s'%self.__class__.__name__)          self.logger = logging.getLogger('inv.%s'%self.__class__.__name__)
104            self.__subsampling_factor=1
"""
is >=1 the value is treated as number of elements to be added to the
domain (per side).
dataset size. For example, calling ``setPadding(3, 0.1)`` to a data
source with size 10x20 will result in the padded data set size
16x24 (10+2*3, 20*(1+2*0.1))

This value is only used for 3-dimensional datasets
"""

def setConstraints(self, bottom=False, sides=True):
"""
If `bottom` is True, then the density mask will be set to 1 in the
padding area at the bottom of the domain. By default this area is
unconstrained. Similarly, if `sides` is True (default) then the
horizontal padding area is constrained, otherwise not.

:param bottom: Whether to constrain the density at the bottom of the
domain
:type bottom: ``bool``
:param sides: Whether to constrain the density in the padding area
surrounding the data
:type sides: ``bool``
"""
self._constrainBottom=bottom
self._constrainSides=sides

def getDomain(self):
"""
Returns a domain that spans the data area plus padding.
The domain is created the first time this method is called, subsequent
calls return the same domain so anything that affects the domain
(such as padding) needs to be set beforehand.

:return: The escript domain for this data source.
:rtype: `esys.escript.Domain`
"""
if self._domain is None:
return self._domain

"""
Returns the density mask data object, where mask has value 1 in the

:return: The mask for the density.
:rtype: `esys.escript.Data`
"""

def getGravityAndStdDev(self):
"""
Returns the gravity anomaly and standard deviation data objects as a
tuple. This method must be implemented in subclasses.
"""
raise NotImplementedError
106
107      def getDataExtents(self):      def getDataExtents(self):
108          """          """
# Line 179  class DataSource(object): Line 116  class DataSource(object):
116          """          """
117          raise NotImplementedError          raise NotImplementedError
118
119      def getVerticalExtents(self):      def getDataType(self):
"""
returns a tuple ``(z0, nz, dz)``, where

- ``z0`` = minimum z coordinate (origin)
- ``nz`` = number of nodes in z direction
- ``dz`` = spacing of nodes (= cell size in z)

This method must be implemented in subclasses.
"""
raise NotImplementedError

def getDomainClass(self):
120          """          """
121          returns the domain generator class (e.g. esys.ripley.Brick).          Returns the type of survey data managed by this source.
122          Must be implemented in subclasses.          Subclasses must return `GRAVITY` or `MAGNETIC` as appropriate.
123          """          """
124          raise NotImplementedError          raise NotImplementedError
125
127          """          """
128          Helper method that computes new number of elements, length and origin          This method is called by the `DomainBuilder` to retrieve the survey
129          after adding padding to the input values.          data as `Data` objects on the given domain.
130            Subclasses should return one or more `Data` objects with survey data
131          :param pad_x: Number of elements or fraction of padding in x direction          interpolated on the given ripley domain. The exact return type
132          :type pad_x: ``int`` or ``float``          depends on the type of data.
133          :param pad_y: Number of elements or fraction of padding in y direction
134          :type pad_y: ``int`` or ``float``          :param domain: the escript domain to use
135          :param NE: Initial number of elements          :type domain: `esys.escript.Domain`
136          :type NE: ``tuple`` or ``list``          :param origin: the origin coordinates of the domain
:param l: Initial side lengths
:type l: ``tuple`` or ``list``
:param origin: Initial origin
137          :type origin: ``tuple`` or ``list``          :type origin: ``tuple`` or ``list``
138          :return: tuple with three elements ``(NE_padded, l_padded, origin_padded)``,          :param NE: the number of domain elements in each dimension
139                   which are lists of the updated input parameters          :type NE: ``tuple`` or ``list``
140          """          :param spacing: the cell sizes (node spacing) in the domain
141          DIM=len(NE)          :type spacing: ``tuple`` or ``list``
frac=[0.]*(DIM-1)+[0]
# padding is applied to each side so multiply by 2 to get the total
# amount of padding per dimension
if DIM>2:

# calculate new number of elements
NE_new=[int(NE[i]*(1+frac[i])) for i in xrange(DIM)]
NEdiff=[NE_new[i]-NE[i] for i in xrange(DIM)]
spacing=[l[i]/NE[i] for i in xrange(DIM)]
l_new=[NE_new[i]*spacing[i] for i in xrange(DIM)]
origin_new=[origin[i]-NEdiff[i]/2.*spacing[i] for i in xrange(DIM)]
return NE_new, l_new, origin_new

def _interpolateOnDomain(self, data):
"""
Helper method that interpolates data arrays onto the domain.
Currently this works like a nearest neighbour mapping, i.e. values
are directly inserted into data objects at closest location.
"""
dom=self.getDomain()
dim=dom.getDim()
# determine number of values required per element
DPP=Scalar(0., ReducedFunction(dom)).getNumberOfDataPoints()
for i in xrange(dim):
DPP=DPP/self._dom_NE[i]
DPP=int(DPP)

# idx_mult.dot([x,y,z]) = flat index into data object
idx_mult=np.array([DPP]+self._dom_NE[:dim-1]).cumprod()

# separate data arrays and coordinates
num_arrays=len(data[0])-dim
arrays=[]
for i in xrange(num_arrays):
d=Scalar(0., ReducedFunction(dom))
d.expand()
arrays.append(d)

for entry in data:
index=[int((entry[i]-self._dom_origin[i])/self._spacing[i]) for i in xrange(dim)]
index=int(idx_mult.dot(index))
for i in xrange(num_arrays):
for p in xrange(DPP):
arrays[i].setValueOfDataPoint(index+p, entry[dim+i])

return arrays

"""
Creates and returns an escript domain that spans the entire area of
available data plus a buffer zone. This method is called only once
the first time `getDomain()` is invoked and may be overwritten if
required.

:return: The escript domain for this data source.
:rtype: `esys.escript.Domain`
"""
X0, NX, DX = self.getDataExtents()
z0, nz, dz = self.getVerticalExtents()

# number of elements (without padding)
NE = [NX[0], NX[1], nz]

# origin of domain (without padding)
origin = [X0[0], X0[1], z0]
origin = [np.round(oi) for oi in origin]

# cell size / point spacing
self._spacing = DX+[dz]
self._spacing = [float(np.round(si)) for si in self._spacing]

# length of domain (without padding)
l = [NE[i]*self._spacing[i] for i in xrange(len(NE))]

NE, l, origin)

# number of padding elements per side

self._dom_len = l_new
self._dom_NE = NE_new
self._dom_origin = origin_new
lo=[(origin_new[i], origin_new[i]+l_new[i]) for i in xrange(3)]
try:
dom=self.getDomainClass()(*self._dom_NE, l0=lo[0], l1=lo[1], l2=lo[2])
# ripley may internally adjust NE and length, so recompute
self._dom_len=[sup(dom.getX()[i])-inf(dom.getX()[i]) for i in xrange(3)]
self._dom_NE=[int(self._dom_len[i]/self._spacing[i]) for i in xrange(3)]

except TypeError:
dom=self.getDomainClass()(*self._dom_NE, l0=l_new[0], l1=l_new[1], l2=l_new[2])

if self._constrainSides:
for i in xrange(2):

if self._constrainBottom:

self.logger.debug("Domain size: %d x %d x %d elements"%(self._dom_NE[0],self._dom_NE[1],self._dom_NE[2]))
self.logger.debug("     length: %g x %g x %g"%(self._dom_len[0],self._dom_len[1],self._dom_len[2]))
self.logger.debug("     origin: %g x %g x %g"%(origin_new[0],origin_new[1],origin_new[2]))

return dom

##############################################################################
class UBCDataSource(DataSource):
def __init__(self, domainclass, meshfile, gravfile, topofile=None):
super(UBCDataSource,self).__init__()
self.__meshfile=meshfile
self.__gravfile=gravfile
self.__topofile=topofile
self.__domainclass=domainclass

numDataPoints=meshdata[0].split()
origin=meshdata[1].split()
self.__nPts=map(int, numDataPoints)
self.__origin=map(float, origin)
self.__delta=[float(X.split('*')[1]) for X in meshdata[2:]]
# vertical data is upside down
self.__origin[2]-=(self.__nPts[2]-1)*self.__delta[2]
self.logger.debug("Data Source: %s (mesh file: %s)"%(self.__gravfile, self.__meshfile))

def getDataExtents(self):
"""
returns ( (x0, y0), (nx, ny), (dx, dy) )
"""
return (self.__origin[:2], self.__nPts[:2], self.__delta[:2])

def getVerticalExtents(self):
"""
returns (z0, nz, dz)
"""
return (self.__origin[2], self.__nPts[2], self.__delta[2])

def getDomainClass(self):
"""
returns the domain generator class (e.g. esys.ripley.Brick)
"""
return self.__domainclass

def getGravityAndStdDev(self):
g_and_sigma=self._interpolateOnDomain(gravlist)
return g_and_sigma[0]*[0,0,1], g_and_sigma[1]

f=open(self.__topofile)
topodata=np.zeros((n,3))
for i in xrange(n):
x=map(float, x)
topodata[i]=x
f.close()

f=open(self.__gravfile)
gravdata=np.zeros((n,5))
for i in xrange(n):
x=map(float, x) # x, y, z, anomaly in mGal, stddev
# convert gravity anomaly units to m/s^2 and rescale error
x[3]*=-1e-5
x[4]*=1e-5
gravdata[i]=x
f.close()
return gravdata

##############################################################################
class NetCDFDataSource(DataSource):
def __init__(self, gravfile, topofile=None, vertical_extents=(-40000,10000,25), alt_of_data=0.):
"""
vertical_extents - (alt_min, alt_max, num_points)
alt_of_data - altitude of measurements
"""
super(NetCDFDataSource,self).__init__()
self.__topofile=topofile
self.__gravfile=gravfile
self.__determineExtents(vertical_extents)
self.__altOfData=alt_of_data

def __determineExtents(self, ve):
self.logger.debug("Data Source: %s"%self.__gravfile)
f=netcdf_file(self.__gravfile, 'r')
NX=0
for n in ['lon','longitude','x']:
if n in f.dimensions:
NX=f.dimensions[n]
break
if NX==0:
raise RuntimeError("Could not determine extents of data")
NY=0
for n in ['lat','latitude','y']:
if n in f.dimensions:
NY=f.dimensions[n]
break
if NY==0:
raise RuntimeError("Could not determine extents of data")

# find longitude and latitude variables
lon_name=None
for n in ['lon','longitude']:
if n in f.variables:
lon_name=n
longitude=f.variables.pop(n)
break
if lon_name is None:
raise RuntimeError("Could not determine longitude variable")
lat_name=None
for n in ['lat','latitude']:
if n in f.variables:
lat_name=n
latitude=f.variables.pop(n)
break
if lat_name is None:
raise RuntimeError("Could not determine latitude variable")

# try to figure out gravity variable name
grav_name=None
if len(f.variables)==1:
grav_name=f.variables.keys()[0]
else:
for n in f.variables.keys():
dims=f.variables[n].dimensions
if (lat_name in dims) and (lon_name in dims):
grav_name=n
break
if grav_name is None:
raise RuntimeError("Could not determine gravity variable")

# try to determine value for unused data
if hasattr(f.variables[grav_name], 'missing_value'):
elif hasattr(f.variables[grav_name], '_FillValue'):
else:

# see if there is a wkt string to convert coordinates
try:
wkt_string=f.variables[grav_name].esri_pe_string
except:
wkt_string=None

# we don't trust actual_range & geospatial_lon_min/max since subset
# data does not seem to have these fields updated.
# Getting min/max from the arrays is obviously not very efficient but..
#lon_range=longitude.actual_range
#lat_range=latitude.actual_range
#lon_range=[f.geospatial_lon_min,f.geospatial_lon_max]
#lat_range=[f.geospatial_lat_min,f.geospatial_lat_max]
lon_range=longitude.data.min(),longitude.data.max()
lat_range=latitude.data.min(),latitude.data.max()
lon_range,lat_range=LatLonToUTM(lon_range, lat_range, wkt_string)
origin=[lon_range[0],lat_range[0],ve[0]]
lengths=[lon_range[1]-lon_range[0], lat_range[1]-lat_range[0],ve[1]-ve[0]]

f.close()

self.__nPts=[NX, NY, ve[2]]
self.__origin=origin
# we are rounding to avoid interpolation issues
self.__delta=[np.round(lengths[i]/self.__nPts[i]) for i in xrange(3)]
self.__wkt_string=wkt_string
self.__lon=lon_name
self.__lat=lat_name
self.__grv=grav_name

def getDataExtents(self):
"""
returns ( (x0, y0), (nx, ny), (dx, dy) )
142          """          """
143          return (self.__origin[:2], self.__nPts[:2], self.__delta[:2])          raise NotImplementedError
144
145      def getVerticalExtents(self):      def setSubsamplingFactor(self, f):
146          """          """
147          returns (z0, nz, dz)          Sets the data subsampling factor (default=1).
148            The factor is applied in all dimensions. For example a 2D dataset
149            with 300 x 150 data points will be reduced to 150 x 75 when a
150            subsampling factor of 2 is used.
151            This becomes important when adding data of varying resolution to
152            a `DomainBuilder`.
153          """          """
154          return (self.__origin[2], self.__nPts[2], self.__delta[2])          self.__subsampling_factor=f
155
156      def getDomainClass(self):      def getSubsamplingFactor(self):
157          """          """
158          returns the domain generator class (e.g. esys.ripley.Brick)          Returns the subsampling factor that was set via `setSubsamplingFactor`
159            (see there).
160          """          """
161          return Brick          return self.__subsampling_factor

def getGravityAndStdDev(self):
nValues=self.__nPts[:2]+[1]
ReducedFunction(self.getDomain()),
g=g*1e-6
sigma=sigma*2e-6
return g*[0,0,1], sigma

f=netcdf_file(self.__topofile, 'r')
lon=None
for n in ['lon','longitude']:
if n in f.variables:
lon=f.variables[n][:]
break
if lon is None:
raise RuntimeError("Could not determine longitude variable")
lat=None
for n in ['lat','latitude']:
if n in f.variables:
lat=f.variables[n][:]
break
if lat is None:
raise RuntimeError("Could not determine latitude variable")
alt=None
for n in ['altitude','alt']:
if n in f.variables:
alt=f.variables[n][:]
break
if alt is None:
raise RuntimeError("Could not determine altitude variable")
162
topodata=np.column_stack((lon,lat,alt))
f.close()
163
164  ##############################################################################  ##############################################################################
165  class ERSDataSource(DataSource):  class ErMapperData(DataSource):
166      """      """
167      Data Source for ER Mapper raster data.      Data Source for ER Mapper raster data.
168      Note that this class only accepts a very specific type of ER Mapper data      Note that this class only accepts a very specific type of ER Mapper data
169      input and will raise an exception if other data is found.      input and will raise an exception if other data is found.
170      """      """
171      def __init__(self, headerfile, datafile=None, vertical_extents=(-40000,10000,25), alt_of_data=0.):      def __init__(self, datatype, headerfile, datafile=None, altitude=0.):
172          """          """
173          headerfile - usually ends in .ers          :param datatype: type of data, must be `GRAVITY` or `MAGNETIC`
174          datafile - usually has the same name as the headerfile without '.ers'          :type datatype: ``int``
177            :param datafile: ER Mapper binary data file name. If not supplied the
178                             name of the header file without '.ers' is assumed
179            :type datafile: ``str``
180            :param altitude: altitude of measurements above ground in meters
181            :type altitude: ``float``
182          """          """
183          super(ERSDataSource,self).__init__()          super(ErMapperData,self).__init__()
185          if datafile is None:          if datafile is None:
187          else:          else:
188              self.__datafile=datafile              self.__datafile=datafile
190          self.__altOfData=alt_of_data          self.__datatype=datatype
192
197          start=-1          start=-1
200                  start=i+1                  start=i+1
201          if start==-1:          if start==-1:
# Line 609  class ERSDataSource(DataSource): Line 203  class ERSDataSource(DataSource):
203
204          md_dict={}          md_dict={}
205          section=[]          section=[]
208              if line[-6:].strip() == 'Begin':              if line[-6:].strip() == 'Begin':
209                  section.append(line[:-6].strip())                  section.append(line[:-6].strip())
# Line 681  class ERSDataSource(DataSource): Line 275  class ERSDataSource(DataSource):
275          # data sets have origin in top-left corner so y runs top-down          # data sets have origin in top-left corner so y runs top-down
276          self.__dataorigin=[originX, originY]          self.__dataorigin=[originX, originY]
277          originY-=(NY-1)*spacingY          originY-=(NY-1)*spacingY
280          self.__delta = [spacingX, spacingY, spacingZ]          self.__nPts = [NX, NY]
281          self.__nPts = [NX, NY, ve[2]]          self.__origin = [originX, originY]
282          self.__origin = [originX, originY, ve[0]]          if self.__datatype == self.GRAVITY:
283                self.logger.info("Assuming gravity data scale is 1e-6 m/s^2.")
284                self.__scalefactor = 1e-6
285            else:
286                self.logger.info("Assuming magnetic data units are 'nT'.")
287                self.__scalefactor = 1e-9
288
289      def getDataExtents(self):      def getDataExtents(self):
290          """          """
291          returns ( (x0, y0), (nx, ny), (dx, dy) )          returns ( (x0, y0), (nx, ny), (dx, dy) )
292          """          """
293          return (self.__origin[:2], self.__nPts[:2], self.__delta[:2])          return (list(self.__origin), list(self.__nPts), list(self.__delta))
294
295      def getVerticalExtents(self):      def getDataType(self):
296          """          return self.__datatype
297          returns (z0, nz, dz)
298        def getSurveyData(self, domain, origin, NE, spacing):
299            nValues=self.__nPts
300            # determine base location of this dataset within the domain
301            first=[int((self.__origin[i]-origin[i])/spacing[i]) for i in range(len(self.__nPts))]
302            if domain.getDim()==3:
303                first.append(int((self.__altitude-origin[2])/spacing[2]))
304                nValues=nValues+[1]
305
307                    ReducedFunction(domain),
310            data = data*self.__scalefactor
311            sigma = sigma * 2. * self.__scalefactor
312            return data, sigma
313
314
315    ##############################################################################
316    class NetCdfData(DataSource):
317        """
318        Data Source for gridded netCDF data that use CF/COARDS conventions.
319        """
320        def __init__(self, datatype, filename, altitude=0.):
321          """          """
322          return (self.__origin[2], self.__nPts[2], self.__delta[2])          :param filename: file name for survey data in netCDF format
323            :type filename: ``str``
324            :param datatype: type of data, must be `GRAVITY` or `MAGNETIC`
325            :type datatype: ``int``
326            :param altitude: altitude of measurements in meters
327            :type altitude: ``float``
328            """
329            super(NetCdfData,self).__init__()
330            self.__filename=filename
331            if not datatype in [self.GRAVITY,self.MAGNETIC]:
332                raise ValueError("Invalid value for datatype parameter")
333            self.__datatype=datatype
334            self.__altitude=altitude
336
338            self.logger.debug("Checking Data Source: %s"%self.__filename)
339            f=netcdf_file(self.__filename, 'r')
340            NX=0
341            for n in ['lon','longitude','x']:
342                if n in f.dimensions:
343                    NX=f.dimensions[n]
344                    break
345            if NX==0:
346                raise RuntimeError("Could not determine extents of data")
347            NY=0
348            for n in ['lat','latitude','y']:
349                if n in f.dimensions:
350                    NY=f.dimensions[n]
351                    break
352            if NY==0:
353                raise RuntimeError("Could not determine extents of data")
354
355      def getDomainClass(self):          # find longitude and latitude variables
356            lon_name=None
357            for n in ['lon','longitude']:
358                if n in f.variables:
359                    lon_name=n
360                    longitude=f.variables.pop(n)
361                    break
362            if lon_name is None:
363                raise RuntimeError("Could not determine longitude variable")
364            lat_name=None
365            for n in ['lat','latitude']:
366                if n in f.variables:
367                    lat_name=n
368                    latitude=f.variables.pop(n)
369                    break
370            if lat_name is None:
371                raise RuntimeError("Could not determine latitude variable")
372
373            # try to figure out data variable name
374            data_name=None
375            if len(f.variables)==1:
376                data_name=f.variables.keys()[0]
377            else:
378                for n in f.variables.keys():
379                    dims=f.variables[n].dimensions
380                    if (lat_name in dims) and (lon_name in dims):
381                        data_name=n
382                        break
383            if data_name is None:
384                raise RuntimeError("Could not determine data variable")
385
386            # try to determine value for unused data
387            if hasattr(f.variables[data_name], 'missing_value'):
389            elif hasattr(f.variables[data_name], '_FillValue'):
391            else:
394
395            # try to determine units of data - this is disabled for now
396            #if hasattr(f.variables[data_name], 'units'):
397            #   units=f.variables[data_name].units
398            if self.__datatype == self.GRAVITY:
399                self.logger.info("Assuming gravity data scale is 1e-6 m/s^2.")
400                self.__scalefactor = 1e-6
401            else:
402                self.logger.info("Assuming magnetic data units are 'nT'.")
403                self.__scalefactor = 1e-9
404
405            # see if there is a wkt string to convert coordinates
406            try:
407                wkt_string=f.variables[data_name].esri_pe_string
408            except:
409                wkt_string=None
410
411            # we don't trust actual_range & geospatial_lon_min/max since subset
412            # data does not seem to have these fields updated.
413            # Getting min/max from the arrays is obviously not very efficient but..
414            #lon_range=longitude.actual_range
415            #lat_range=latitude.actual_range
416            #lon_range=[f.geospatial_lon_min,f.geospatial_lon_max]
417            #lat_range=[f.geospatial_lat_min,f.geospatial_lat_max]
418            lon_range=longitude.data.min(),longitude.data.max()
419            lat_range=latitude.data.min(),latitude.data.max()
420            if lon_range[1]<180:
421                lon_range,lat_range=LatLonToUTM(lon_range, lat_range, wkt_string)
422            lengths=[lon_range[1]-lon_range[0], lat_range[1]-lat_range[0]]
423            f.close()
424
425            self.__nPts=[NX, NY]
426            self.__origin=[lon_range[0],lat_range[0]]
427            # we are rounding to avoid interpolation issues
428            self.__delta=[np.round(lengths[i]/self.__nPts[i]) for i in range(2)]
429            #self.__wkt_string=wkt_string
430            #self.__lon_name=lon_name
431            #self.__lat_name=lat_name
432            self.__data_name=data_name
434
435        def getDataExtents(self):
436          """          """
437          returns the domain generator class (e.g. esys.ripley.Brick)          returns ( (x0, y0), (nx, ny), (dx, dy) )
438          """          """
439          return Brick          return (list(self.__origin), list(self.__nPts), list(self.__delta))
440
441      def getGravityAndStdDev(self):      def getDataType(self):
442          nValues=self.__nPts[:2]+[1]          return self.__datatype
444          g=ripleycpp._readBinaryGrid(self.__datafile,      def getSurveyData(self, domain, origin, NE, spacing):
445                  ReducedFunction(self.getDomain()),          nValues=self.__nPts
446                  first, nValues, (), self.__maskval)          # determine base location of this dataset within the domain
447          sigma=whereNonZero(g-self.__maskval)          first=[int((self.__origin[i]-origin[i])/spacing[i]) for i in range(len(self.__nPts))]
448          g=g*1e-6          if domain.getDim()==3:
449          sigma=sigma*2e-6              first.append(int((self.__altitude-origin[2])/spacing[2]))
450          return g*[0,0,1], sigma              nValues=nValues+[1]
451
453                      ReducedFunction(domain), first, nValues, (), self.__maskval)
455            data=data*self.__scalefactor
456            sigma=sigma * 2. * self.__scalefactor
457            return data, sigma
458
459
460  ##############################################################################  ##############################################################################
# Line 724  class SourceFeature(object): Line 464  class SourceFeature(object):
464      data source, for example a layer of a specific rock type or a simulated      data source, for example a layer of a specific rock type or a simulated
465      ore body.      ore body.
466      """      """
467      def getDensity(self):      def getValue(self):
468          """          """
469          Returns the density for the area covered by mask. It can be constant          Returns the value for the area covered by mask. It can be constant
470          or a data object with spatial dependency.          or a data object with spatial dependency.
471          """          """
472          raise NotImplementedError          raise NotImplementedError
# Line 739  class SourceFeature(object): Line 479  class SourceFeature(object):
479          raise NotImplementedError          raise NotImplementedError
480
481  class SmoothAnomaly(SourceFeature):  class SmoothAnomaly(SourceFeature):
482      def __init__(self, lx, ly, lz, x, y, depth, rho_inner, rho_outer):      def __init__(self, lx, ly, lz, x, y, depth, v_inner=None, v_outer=None):
483          self.x=x          self.x=x
484          self.y=y          self.y=y
485          self.lx=lx          self.lx=lx
486          self.ly=ly          self.ly=ly
487          self.lz=lz          self.lz=lz
488          self.depth=depth          self.depth=depth
489          self.rho_inner=rho_inner          self.v_inner=v_inner
490          self.rho_outer=rho_outer          self.v_outer=v_outer
491          self.rho=None          self.value=None
493      def getDensity(self):
494          return self.rho      def getValue(self,x):
495            if self.value is None:
496                if self.v_outer is None or self.v_inner is None:
497                    self.value=0
498                else:
499                    DIM=x.getDomain().getDim()
500                    alpha=-log(abs(self.v_outer/self.v_inner))*4
501                    value=exp(-alpha*((x[0]-self.x)/self.lx)**2)
502                    value=value*exp(-alpha*((x[DIM-1]+self.depth)/self.lz)**2)
503                    self.value=maximum(abs(self.v_outer), abs(self.v_inner*value))
504                    if self.v_inner<0: self.value=-self.value
505
506            return self.value
507
509          DIM=x.getDomain().getDim()          DIM=x.getDomain().getDim()
510          m=whereNonNegative(x[DIM-1]-(sup(x[DIM-1])-self.depth-self.lz/2)) * whereNonPositive(x[DIM-1]-(sup(x[DIM-1])-self.depth+self.lz/2)) \          m=whereNonNegative(x[DIM-1]+self.depth+self.lz/2) * whereNonPositive(x[DIM-1]+self.depth-self.lz/2) \
511              *whereNonNegative(x[0]-(self.x-self.lx/2)) * whereNonPositive(x[0]-(self.x+self.lx/2))              *whereNonNegative(x[0]-(self.x-self.lx/2)) * whereNonPositive(x[0]-(self.x+self.lx/2))
512          if DIM>2:          if DIM>2:
513              m*=whereNonNegative(x[1]-(self.y-self.ly/2)) * whereNonPositive(x[1]-(self.y+self.ly/2))              m*=whereNonNegative(x[1]-(self.y-self.ly/2)) * whereNonPositive(x[1]-(self.y+self.ly/2))
514          if self.rho is None:          self.mask = m
515              alpha=-log(abs(self.rho_outer/self.rho_inner))*4          return m
rho=exp(-alpha*((x[0]-self.x)/self.lx)**2)
rho=rho*exp(-alpha*((x[DIM-1]-(sup(x[DIM-1])-self.depth))/self.lz)**2)
self.rho=maximum(abs(self.rho_outer), abs(self.rho_inner*rho))
if self.rho_inner<0: self.rho=-self.rho
return m
516
517  ##############################################################################  ##############################################################################
518  class SyntheticDataSource(DataSource):  class SyntheticFeatureData(DataSource):
519      def __init__(self, DIM, NE, l, h, features):      def __init__(self, datatype, DIM, NE, l, features, B_b=None):
520          super(SyntheticDataSource,self).__init__()          super(SyntheticFeatureData,self).__init__()
521            if not datatype in [self.GRAVITY,self.MAGNETIC]:
522                raise ValueError("Invalid value for datatype parameter")
523            self.__datatype = datatype
524          self._features = features          self._features = features
525            self.__origin = [0]*(DIM-1)
526            self.__nPts = [NE]*(DIM-1)
527            self.__delta = [float(l)/NE]*(DIM-1)
528            self.__B_b =None
529          self.DIM=DIM          self.DIM=DIM
530          self.NE=NE          self.NE=NE
531          self.l=l          self.l=l
532          self.h=h          # this is for Cartesian (FIXME ?)
533            if datatype  ==  self.MAGNETIC:
535          NE_H=self.NE                 self.__B_b =  np.array([-B_b[2],  -B_b[0]])
536          NE_L=int((self.l/self.h)*NE_H+0.5)              else:
537          l=[self.l]*(self.DIM-1)+[self.h]                 self.__B_b = ([-B_b[1],  -B_b[2],  -B_b[0]])
NE=[NE_L]*(self.DIM-1)+[NE_H]
origin=[0.]*self.DIM
NE, l, origin)

self.NE=NE_new
self.l=l_new[0]
self.h=l_new[self.DIM-1]

self.logger.debug("Data Source: synthetic with %d features"%len(self._features))
if self.DIM==2:
from esys.finley import Rectangle
dom = Rectangle(n0=NE_new[0], n1=NE_new[1], l0=l_new[0], l1=l_new[1])
self._x = dom.getX() + origin_new
self.logger.debug("Domain size: %d x %d elements"%(NE_new[0], NE_new[1]))
self.logger.debug("     length: %g x %g"%(l_new[0],l_new[1]))
self.logger.debug("     origin: %g x %g"%(origin_new[0],origin_new[1]))
else:
from esys.finley import Brick
dom = Brick(n0=NE_new[0], n1=NE_new[1], n2=NE_new[2], l0=l_new[0], l1=l_new[1], l2=l_new[2])
self._x = dom.getX() + origin_new
self.logger.debug("Domain size: %d x %d x %d elements"%(self.NE[0],self.NE[1],self.NE[2]))
self.logger.debug("     length: %g x %g x %g"%(l_new[0],l_new[1],l_new[2]))
self.logger.debug("     origin: %g x %g x %g"%(origin_new[0],origin_new[1],origin_new[2]))

dz=l_new[self.DIM-1]/NE_new[self.DIM-1]
* whereNegative(dom.getX()[0]-(l_new[0]-origin_new[0])) \
* whereNonNegative(dom.getX()[self.DIM-1]-(l_new[self.DIM-1]+origin_new[self.DIM-1])) \
* whereNonPositive(dom.getX()[self.DIM-1]-(l_new[self.DIM-1]+(origin_new[self.DIM-1]+dz)))
wherePositive(self._x[self.DIM-1]-l[self.DIM-1])
for i in xrange(self.DIM-1):
wherePositive(self._x[i]-l[i])

rho_ref=0.
for f in self._features:
rho_ref = rho_ref * (1-m) + f.getDensity() * m
self._rho=rho_ref
538
539          return dom      def getDataExtents(self):
540            return (list(self.__origin), list(self.__nPts), list(self.__delta))
541
544
545      def getReferenceDensity(self):      def getReferenceDensity(self):
546            """
547            Returns the reference density Data object that was used to generate
548            the gravity anomaly data.
549            """
550          return self._rho          return self._rho
551
552      def getGravityAndStdDev(self):      def getReferenceSusceptibility(self):
553          pde=LinearSinglePDE(self.getDomain())          """
554          G=6.6742e-11*U.m**3/(U.kg*U.sec**2)          Returns the reference magnetic susceptibility Data objects that was
555          m_psi_ref=0.          used to generate the magnetic field data.
556          for i in xrange(self.DIM):          """
557              m_psi_ref=m_psi_ref + whereZero(self._x[i]-inf(self._x[i])) \          return self._k
+ whereZero(self._x[i]-sup(self._x[i]))
558
559          pde.setValue(A=kronecker(self.getDomain()), Y=-4*np.pi*G*self._rho, q=m_psi_ref)      def getSurveyData(self, domain, origin, NE, spacing):
560            pde=LinearSinglePDE(domain)
561            G=U.Gravitational_Constant
562            m_psi_ref=0.
563            x=domain.getX()
564            DIM=domain.getDim()
565            m_psi_ref=whereZero(x[DIM-1]-sup(x[DIM-1])) # + whereZero(x[DIM-1]-inf(x[DIM-1]))
566            if self.getDataType()==DataSource.GRAVITY:
567                rho_ref=0.
568                for f in self._features:
570                    rho_ref = rho_ref * (1-m) + f.getValue(x) * m
571                self._rho=rho_ref
572                pde.setValue(A=kronecker(domain), Y=-4*np.pi*G*rho_ref, q=m_psi_ref)
573            else:
574                k_ref=0.
575                for f in self._features:
577                    k_ref = k_ref * (1-m) + f.getValue(x) * m
578                self._k=k_ref
579                pde.setValue(A=kronecker(domain), X=k_ref*self.__B_b, q=m_psi_ref)
580          pde.setSymmetryOn()          pde.setSymmetryOn()
581          psi_ref=pde.getSolution()          psi_ref=pde.getSolution()
582          del pde          del pde
585          return g,sigma          else:
587
588            sigma=1.
590            x=ReducedFunction(domain).getX()
591            for i in range(self.DIM-1):
592                sigma=sigma * wherePositive(x[i]) \
593                            * whereNegative(x[i]-(sup(x[i])+inf(x[i])))
594            # limit mask to one cell thickness at z=0
595            sigma = sigma * whereNonNegative(x[self.DIM-1]) \
596                    * whereNonPositive(x[self.DIM-1]-spacing[self.DIM-1])
597            return data,sigma
598
599
600

Legend:
 Removed from v.4019 changed lines Added in v.4108