class SparseMatrix:
def __init__(self, dim, defaultVal = 0):
self.dim = dim
self.defVal = defaultVal
self.trunk = TTTree()
self.getidx = []
self.maxIdx = self.__indexFunc(map(lambda x: x - 1, self.dim))
def __indexFunc(self, idx):
# returns linear index by the given n-d index
linearIdx = idx[0]
if len(idx) > 1:
for i in xrange(1, len(self.dim)):
linearIdx += reduce(lambda x, y: x * y, self.dim[:i]) * idx[i]
# check bounds
if 'maxIdx' in self.__dict__ and linearIdx > self.maxIdx:
print 'Index out of bounds!'
return linearIdx
def __invIndexFunc(self, idx):
# returns n-d index by the given linear index
output = [0] * len(self.dim)
for i in xrange(len(self.dim) - 1, 0, -1):
divisor = reduce(lambda x, y: x * y, self.dim[:i])
result = idx / divisor
idx -= result * divisor
output[i] = result
output[0] = idx
return output
def __getitem__(self, idx):
self.getidx.append(idx)
if len(self.getidx) == len(self.dim):
linidx = self.__indexFunc(self.getidx)
node = self.trunk.contains(linidx)
self.getidx = []
if node:
return node.getItem(linidx).val()
return self.defVal
return self
def setitem(self, idxlst, val):
if len(idxlst) == len(self.dim):
linidx = self.__indexFunc(idxlst)
res = self.trunk.insertValue(Pair(linidx, val))
if type(res) is Pair: res.value = val
return self
def size(self):
return self.dim
def __str__(self):
buf = []
for node in self.trunk:
for pair in node:
if pair is not None:
buf.append('%s: %s' % (self.__invIndexFunc(pair.key), pair.value))
return ', '.join(buf)
class SparseMatrix:
def __init__(self, dim, defaultVal=0):
self.dim = dim
self.defVal = defaultVal
self.trunk = TTTree()
self.getidx = []
self.maxIdx = self.__indexFunc(map(lambda x: x - 1, self.dim))
def __indexFunc(self, idx):
# returns linear index by the given n-d index
linearIdx = idx[0]
if len(idx) > 1:
for i in xrange(1, len(self.dim)):
linearIdx += reduce(lambda x, y: x * y, self.dim[:i]) * idx[i]
# check bounds
if 'maxIdx' in self.__dict__ and linearIdx > self.maxIdx:
print 'Index out of bounds!'
return linearIdx
def __invIndexFunc(self, idx):
# returns n-d index by the given linear index
output = [0] * len(self.dim)
for i in xrange(len(self.dim) - 1, 0, -1):
divisor = reduce(lambda x, y: x * y, self.dim[:i])
result = idx / divisor
idx -= result * divisor
output[i] = result
output[0] = idx
return output
def __getitem__(self, idx):
self.getidx.append(idx)
if len(self.getidx) == len(self.dim):
linidx = self.__indexFunc(self.getidx)
node = self.trunk.contains(linidx)
self.getidx = []
if node:
return node.getItem(linidx).val()
return self.defVal
return self
def setitem(self, idxlst, val):
if len(idxlst) == len(self.dim):
linidx = self.__indexFunc(idxlst)
res = self.trunk.insertValue(Pair(linidx, val))
if type(res) is Pair: res.value = val
return self
def size(self):
return self.dim
def __str__(self):
buf = []
for node in self.trunk:
for pair in node:
if pair is not None:
buf.append('%s: %s' %
(self.__invIndexFunc(pair.key), pair.value))
return ', '.join(buf)
