def setDF(self, df, update=1):
# config = self.get_config()
configDict = self.__getstate__()
[configDict.pop(k) for k in list(configDict) if k.startswith('_')]
if isinstance(df, pd.Series):
df = df.values
if isinstance(df, pyutil.np.ndarray):
if update and (df.shape == self.shape):
self.loc[:, :] = df
res = self
else:
res = pyutil.init_DF(df, rowName=self.index)
else:
res = df
if not isinstance(res, self.__class__):
assert isinstance(res, pd.DataFrame)
# print self.__class__, res.__class__
res = self.__class__.from_DataFrame(df=res, )
res.__setstate__(configDict)
res.set__colMeta(res.colMeta)
res.set__rowMeta(res.rowMeta)
# res.set_config(**config)
return res
def __init__(self,
C=None,
rowName=None,
colName=None,
name=None,
look=None,
cmap=None,
vlim=None,
fname=None,
model=None,
**kwargs):
df = pyutil.init_DF(C=C, rowName=rowName, colName=colName)
# print self.__class__,countMatrix, isinstance(self,countMatrix,)
super(countMatrix, self).__init__(df)
self.name_ = name
self.look = look
self.cmap = cmap
self.vlim = vlim
self.fname = fname
self.model = model
self.param = {
'normF': 'identityNorm',
}
self.set_config(test=None, **kwargs)
self.test = None
def __init__(self,
C=None,
rowName=None,
colName=None,
name=None,
look=None,
cmap=None,
vlim=None,
fname=None,
model=None,
colMeta=None,
rowMeta=None,
height=1.,
**kwargs):
df = pyutil.init_DF(C=C, rowName=rowName, colName=colName)
# print self.__class__,countMatrix, isinstance(self,countMatrix,)
super(countMatrix, self).__init__(df)
self.name_ = name
self.look = look
self.cmap = cmap
self.vlim = vlim
self.fname = fname
self.model = model
self.height = height
# self.colMeta_ = colMeta
self.set__colMeta(colMeta)
self.set__rowMeta(rowMeta)
# self.rowMeta_ = rowMeta
self.param = {
'normF': 'identityNorm',
}
self.set_config(test=None, **kwargs)
self.test = None
assert self.name != 'test', 'Track name cannot be %s' % self.name
def from_deepcache(cls, d=None, fname=None):
if fname is not None:
d = np.load(fname)
d = dict(d)
else:
assert d is not None, 'must specify one of variables: "d" or "fname"'
deepdict = d
C = deepdict.get('matrix', np.array([[]]))
colName = deepdict.get('labels', None)
rowName = deepdict.get('which is row name?', None)
df = pyutil.init_DF(C=C, colName=colName, rowName=rowName)
ins = cls.from_DataFrame(df=df)
return ins
def setDF(self, df):
config = self.get_config()
if isinstance(df, pd.Series):
df = df.values
if isinstance(df, pyutil.np.ndarray):
if df.shape == self.shape:
self.loc[:, :] = df
res = self
else:
res = pyutil.init_DF(df, rowName=self.index)
else:
res = df
if not isinstance(res, self.__class__):
assert isinstance(res, pd.DataFrame)
# print self.__class__, res.__class__
res = self.__class__.from_DataFrame(df=res, )
res.set_config(**config)
return res