def filter(self, boolarray):
if boolarray.squeeze().ndim > 1:
raise ValueError("filter can only use array with one dimension")
if boolarray.dtype != np.dtype(bool):
if boolarray.dims[0].name not in self.ivardata:
out = self.copy()
msg = "Filter warning: DataBlock does not contain dimension %r"
msg = msg % boolarray.dims[0]
warn(msg)
return out
localdim = self.ivardata[boolarray.dims[0].name]
intersection = set(localdim.data).intersection(boolarray)
barray = boolarray.astype(bool)
barray[...] = False
for f in intersection:
barray[boolarray == f] = True
boolarray = barray
out = DataBlock()
out.blockname = self.blockname
out.comments = self.comments
out.xname = self.xname
olddim = boolarray.dims[0]
newdim = olddim.__class__(olddim, data=olddim.data[boolarray])
for v in self.vardata.keys():
data = self.vardata[v].view()
try:
i = data.dims_index(boolarray.dims[0])
except IndexError: # This variable does not sweep in boolarray dim
out[v] = data
continue
newinfo = list(data.dims)
reorder = list(range(data.ndim))
del newinfo[i]
del reorder[i]
reorder = tuple([i] + reorder)
newinfo = tuple([newdim] + newinfo)
data = data.transpose(*reorder)
utdata = hfarray(array(data)[boolarray],
dims=newinfo,
unit=data.unit,
outputformat=data.outputformat)
dimorder = [x.name for x in self.vardata[v].dims]
utdata = utdata.reorder_dimensions(*dimorder)
out[v] = utdata
out.xname = self.xname
return out
def filter(self, boolarray):
if boolarray.squeeze().ndim > 1:
raise ValueError("filter can only use array with one dimension")
if boolarray.dtype != np.dtype(bool):
if boolarray.dims[0].name not in self.ivardata:
out = self.copy()
msg = "Filter warning: DataBlock does not contain dimension %r"
msg = msg % boolarray.dims[0]
warn(msg)
return out
localdim = self.ivardata[boolarray.dims[0].name]
intersection = set(localdim.data).intersection(boolarray)
barray = boolarray.astype(bool)
barray[...] = False
for f in intersection:
barray[boolarray == f] = True
boolarray = barray
out = DataBlock()
out.blockname = self.blockname
out.comments = self.comments
out.xname = self.xname
olddim = boolarray.dims[0]
newdim = olddim.__class__(olddim, data=olddim.data[boolarray])
for v in self.vardata.keys():
data = self.vardata[v].view()
try:
i = data.dims_index(boolarray.dims[0])
except IndexError: # This variable does not sweep in boolarray dim
out[v] = data
continue
newinfo = list(data.dims)
reorder = list(range(data.ndim))
del newinfo[i]
del reorder[i]
reorder = tuple([i] + reorder)
newinfo = tuple([newdim] + newinfo)
data = data.transpose(*reorder)
utdata = hfarray(array(data)[boolarray], dims=newinfo,
unit=data.unit,
outputformat=data.outputformat)
dimorder = [x.name for x in self.vardata[v].dims]
utdata = utdata.reorder_dimensions(*dimorder)
out[v] = utdata
out.xname = self.xname
return out
def __new__(subtype,
data,
dims=None,
dtype=None,
copy=True,
order=None,
subok=False,
ndmin=0,
unit=None,
info=None,
Z0=None):
if info is not None:
warn("_MultiPortArray, use dims not info")
if dims is not None:
raise ValueError("Can not specify both dims and info")
dims = info
info = None
data = np.asanyarray(data)
if dims is None:
if hasattr(data, 'dims'):
dims = tuple(data.dims)
elif len(data.shape) <= len(subtype.default_dim) + 2:
defaultdims = (subtype.default_dim)
shape = data.shape[:-2]
dims = tuple(
x.__class__(x.name, range(size))
for (x, size) in zip(defaultdims, shape))
dims = dims + (DimMatrix_i(
"i", data.shape[-2]), DimMatrix_j("j", data.shape[-1]))
out = _hfarray.__new__(subtype,
data,
dims=dims,
dtype=dtype,
copy=copy,
order=order,
subok=subok,
ndmin=ndmin,
unit=unit)
return out
def sort(self, dim):
dim = hfarray(dim, copy=False).squeeze()
if dim.ndim > 1:
msg = "sort can only get sort direction from array with"\
" one dimension"
raise ValueError(msg)
else:
dimname = dim.dims[0].name
if dimname not in self.ivardata:
msg = "Sort warning: DataBlock does not contain dimension %r"
msg = msg % dimname
warn(msg)
out = self.copy()
return out
sortorder = dim.argsort()
out = DataBlock()
out.blockname = self.blockname
out.comments = self.comments
out.xname = self.xname
for v in self.vardata.keys()[:]:
data = self.vardata[v].view()
try:
i = data.dims_index(dimname)
out[v] = data.take(sortorder, i)
except IndexError: # This variable does not contain sort direction
out[v] = data
continue
olddim = self.ivardata[dimname]
out.xname = self.xname
data = self[dimname][sortorder]
out.replace_dim(olddim, olddim.__class__(olddim, data=data))
return out
def sort(self, dim):
dim = hfarray(dim, copy=False).squeeze()
if dim.ndim > 1:
msg = "sort can only get sort direction from array with"\
" one dimension"
raise ValueError(msg)
else:
dimname = dim.dims[0].name
if dimname not in self.ivardata:
msg = "Sort warning: DataBlock does not contain dimension %r"
msg = msg % dimname
warn(msg)
out = self.copy()
return out
sortorder = dim.argsort()
out = DataBlock()
out.blockname = self.blockname
out.comments = self.comments
out.xname = self.xname
for v in self.vardata.keys()[:]:
data = self.vardata[v].view()
try:
i = data.dims_index(dimname)
out[v] = data.take(sortorder, i)
except IndexError: # This variable does not contain sort direction
out[v] = data
continue
olddim = self.ivardata[dimname]
out.xname = self.xname
data = self[dimname][sortorder]
out.replace_dim(olddim, olddim.__class__(olddim, data=data))
return out
def __new__(subtype, data, dims=None, dtype=None, copy=True, order=None,
subok=False, ndmin=0, unit=None, info=None, Z0=None):
if info is not None:
warn("_MultiPortArray, use dims not info")
if dims is not None:
raise ValueError("Can not specify both dims and info")
dims = info
info = None
data = np.asanyarray(data)
if dims is None:
if hasattr(data, 'dims'):
dims = tuple(data.dims)
elif len(data.shape) <= len(subtype.default_dim) + 2:
defaultdims = (subtype.default_dim)
shape = data.shape[:-2]
dims = tuple(x.__class__(x.name, range(size))
for (x, size) in zip(defaultdims, shape))
dims = dims + (DimMatrix_i("i", data.shape[-2]),
DimMatrix_j("j", data.shape[-1]))
out = _hfarray.__new__(subtype, data, dims=dims, dtype=dtype,
copy=copy, order=order, subok=subok,
ndmin=ndmin, unit=unit)
return out
