def data(self):
"""
Get Selection data as a pandas.DataFrame
If the data is modified, 'save' must be called for CASM to use the modified selection.
"""
if self._data is None:
if self.path in ["MASTER", "ALL", "CALCULATED"]:
self._data = query(self.proj, ['configname', 'selected'],
self,
all=self.all)
elif self._is_json():
self._data = pandas.read_json(self.path, 'r', orient='records')
else:
with open(self.path, compat.pandas_rmode()) as f:
if compat.peek(f) == '#':
f.read(1)
self._data = pandas.read_csv(f,
sep=compat.str(' +'),
engine='python')
self._clean_data()
if not self.all:
self._data = self._data[self._data['selected'] == True]
return self._data
def query(self, columns, force=False, verbose=False):
""" Query requested columns and store them in 'data'.
Will not overwrite columns that already exist, unless 'force'==True. Will query data for all configurations, whether selected or not, if `self.all == True`.
Arguments
---------
columns: List(str)
Data requested, will be added as columns in `self.data`. This corresponds to the `-k` option of `casm query`. A list of options can be obtained from `casm query --help properties`.
force: bool
If `force==False`, input `columns` that already exist in `self.data.columns` will be ignored and those columns will not be updated. If `force==True`, those columns will be overwritten with new data.
verbose: bool
How much to print to stdout.
"""
if force == False:
_col = [x for x in columns if x not in self.data.columns]
else:
_col = columns
if verbose:
print("# Query requested:", columns)
if force == False:
print("# Use existing:",
[x for x in columns if x in self.data.columns])
else:
print("# Overwrite existing:",
[x for x in columns if x in self.data.columns])
if len(_col) == 0:
print("# No query necessary")
else:
print("# Querying:", _col)
if len(_col) == 0:
return
df = query(self.proj,
_col,
self.path,
self.type,
verbatim=True,
all=self.all)
if verbose:
print("# DONE\n")
msg = "querying different numbers of records: {0}, {1}".format(
self.data.shape, df.shape)
assert self.data.shape[0] == df.shape[0], msg
for c in df.columns:
self.data.loc[:, c] = df.loc[:, c].values
def query(self, columns, force=False, verbose=False):
"""
Query requested columns and store them in 'data'. Will not overwrite
columns that already exist, unless 'force'==True.
Will query data for all configurations, whether selected or not, if
self.all == True.
"""
if force == False:
_col = [x for x in columns if x not in self.data.columns]
else:
_col = columns
if verbose:
print "# Query requested:", columns
if force == False:
print "# Use existing:", [
x for x in columns if x in self.data.columns
]
else:
print "# Overwrite existing:", [
x for x in columns if x in self.data.columns
]
if len(_col) == 0:
print "# No query necessary"
else:
print "# Querying:", _col
if len(_col) == 0:
return
df = query(self.proj, _col, self, all=self.all)
if verbose:
print "# DONE\n"
msg = "querying different numbers of records: {0}, {1}".format(
self.data.shape, df.shape)
assert self.data.shape[0] == df.shape[0], msg
for c in df.columns:
self.data.loc[:, c] = df.loc[:, c].values
def data(self):
"""
Get Selection data as a pandas.DataFrame
If the data is modified, 'save' must be called for CASM to use the modified selection.
"""
if self._data is None:
if self.path in ["MASTER", "ALL", "CALCULATED"]:
self._data = query(self.proj, ['configname', 'selected'], self)
elif self._is_json():
self._data = pandas.read_json(self.path, orient='records')
else:
f = open(self.path, 'r')
f.read(1)
self._data = pandas.read_csv(f, sep=' *', engine='python')
self._clean_data()
if not self.all:
self._data = self._data[self._data['selected']==True]
return self._data
def query(self, columns, force=False, verbose=False):
"""
Query requested columns and store them in 'data'. Will not overwrite
columns that already exist, unless 'force'==True.
Will query data for all configurations, whether selected or not, if
self.all == True.
"""
if force == False:
_col = [x for x in columns if x not in self.data.columns]
else:
_col = columns
if verbose:
print "# Query requested:", columns
if force == False:
print "# Use existing:", [x for x in columns if x in self.data.columns]
else:
print "# Overwrite existing:", [x for x in columns if x in self.data.columns]
if len(_col) == 0:
print "# No query necessary"
else:
print "# Querying:", _col
if len(_col) == 0:
return
df = query(self.proj, _col, self, all=self.all)
if verbose:
print "# DONE\n"
msg = "querying different numbers of records: {0}, {1}".format(
self.data.shape, df.shape)
assert self.data.shape[0] == df.shape[0], msg
for c in df.columns:
self.data.loc[:,c] = df.loc[:,c].values
