##// END OF EJS Templates
Merge pull request #2868 from takluyver/import-performance...
Merge pull request #2868 from takluyver/import-performance Import performance Defer various imports for a small reduction in startup time. IPython.lib.__init__ previously loaded IPython.lib.inputhook to export some of its functions. This meant that importing anything from IPython.lib also loaded inputhook. For now, I've just removed that import, but that is an API change, because the functions are no longer accessible as e.g. IPython.lib.enable_qt4 Added a note about the API change. Timing command, borrowed from Openoffice to simulate cold start: $ sync ; echo 3 | sudo tee /proc/sys/vm/drop_caches; time ipython3 -c pass Results with this branch: 7.68, 7.64, 7.35, 7.38 s 'real' Results with master: 8.17, 8.04, 7.81, 7.77

File last commit:

r5560:a3ee8de9
r9459:ff3032f2 merge
Show More
map.py
171 lines | 5.3 KiB | text/x-python | PythonLexer
# encoding: utf-8
"""Classes used in scattering and gathering sequences.
Scattering consists of partitioning a sequence and sending the various
pieces to individual nodes in a cluster.
Authors:
* Brian Granger
* MinRK
"""
#-------------------------------------------------------------------------------
# Copyright (C) 2008-2011 The IPython Development Team
#
# Distributed under the terms of the BSD License. The full license is in
# the file COPYING, distributed as part of this software.
#-------------------------------------------------------------------------------
#-------------------------------------------------------------------------------
# Imports
#-------------------------------------------------------------------------------
from __future__ import division
import types
from itertools import islice
from IPython.utils.data import flatten as utils_flatten
#-------------------------------------------------------------------------------
# Figure out which array packages are present and their array types
#-------------------------------------------------------------------------------
arrayModules = []
try:
import Numeric
except ImportError:
pass
else:
arrayModules.append({'module':Numeric, 'type':Numeric.arraytype})
try:
import numpy
except ImportError:
pass
else:
arrayModules.append({'module':numpy, 'type':numpy.ndarray})
try:
import numarray
except ImportError:
pass
else:
arrayModules.append({'module':numarray,
'type':numarray.numarraycore.NumArray})
class Map:
"""A class for partitioning a sequence using a map."""
def getPartition(self, seq, p, q):
"""Returns the pth partition of q partitions of seq."""
# Test for error conditions here
if p<0 or p>=q:
print "No partition exists."
return
remainder = len(seq)%q
basesize = len(seq)//q
hi = []
lo = []
for n in range(q):
if n < remainder:
lo.append(n * (basesize + 1))
hi.append(lo[-1] + basesize + 1)
else:
lo.append(n*basesize + remainder)
hi.append(lo[-1] + basesize)
try:
result = seq[lo[p]:hi[p]]
except TypeError:
# some objects (iterators) can't be sliced,
# use islice:
result = list(islice(seq, lo[p], hi[p]))
return result
def joinPartitions(self, listOfPartitions):
return self.concatenate(listOfPartitions)
def concatenate(self, listOfPartitions):
testObject = listOfPartitions[0]
# First see if we have a known array type
for m in arrayModules:
#print m
if isinstance(testObject, m['type']):
return m['module'].concatenate(listOfPartitions)
# Next try for Python sequence types
if isinstance(testObject, (types.ListType, types.TupleType)):
return utils_flatten(listOfPartitions)
# If we have scalars, just return listOfPartitions
return listOfPartitions
class RoundRobinMap(Map):
"""Partitions a sequence in a roun robin fashion.
This currently does not work!
"""
def getPartition(self, seq, p, q):
# if not isinstance(seq,(list,tuple)):
# raise NotImplementedError("cannot RR partition type %s"%type(seq))
return seq[p:len(seq):q]
#result = []
#for i in range(p,len(seq),q):
# result.append(seq[i])
#return result
def joinPartitions(self, listOfPartitions):
testObject = listOfPartitions[0]
# First see if we have a known array type
for m in arrayModules:
#print m
if isinstance(testObject, m['type']):
return self.flatten_array(m['type'], listOfPartitions)
if isinstance(testObject, (types.ListType, types.TupleType)):
return self.flatten_list(listOfPartitions)
return listOfPartitions
def flatten_array(self, klass, listOfPartitions):
test = listOfPartitions[0]
shape = list(test.shape)
shape[0] = sum([ p.shape[0] for p in listOfPartitions])
A = klass(shape)
N = shape[0]
q = len(listOfPartitions)
for p,part in enumerate(listOfPartitions):
A[p:N:q] = part
return A
def flatten_list(self, listOfPartitions):
flat = []
for i in range(len(listOfPartitions[0])):
flat.extend([ part[i] for part in listOfPartitions if len(part) > i ])
return flat
#lengths = [len(x) for x in listOfPartitions]
#maxPartitionLength = len(listOfPartitions[0])
#numberOfPartitions = len(listOfPartitions)
#concat = self.concatenate(listOfPartitions)
#totalLength = len(concat)
#result = []
#for i in range(maxPartitionLength):
# result.append(concat[i:totalLength:maxPartitionLength])
# return self.concatenate(listOfPartitions)
def mappable(obj):
"""return whether an object is mappable or not."""
if isinstance(obj, (tuple,list)):
return True
for m in arrayModules:
if isinstance(obj,m['type']):
return True
return False
dists = {'b':Map,'r':RoundRobinMap}