def howlong1(matrixSize,matrixNum):
start = time.time()
for i in range(matrixNum):
X=np.random.random((2*matrixSize,2*matrixSize))
S=X.transpose()-X
pf(S)
end= time.time()
return (end-start)
def track_object(self, obj_i, start_state):
print "Tracking object %i of %i" % (obj_i + 1, len(self.start_states))
print "Start state for object %i is %s." % (obj_i + 1, start_state)
self.preresampled_particles[obj_i] = numpy.zeros((self.num_frames, self.num_particles, start_state.size))
self.resampled_particles[obj_i] = numpy.zeros((self.num_frames, self.num_particles, start_state.size))
self.highest_weight_particles.append([start_state])
particles = numpy.array([start_state] * self.num_particles)
track = numpy.zeros((self.num_frames, start_state.size))
track[0, :] = start_state
self.preresampled_particles[obj_i][0] = particles
self.resampled_particles[obj_i][0] = particles
for i, frame in enumerate(self.video[1:], 1):
particles, intermediate_particles = pf(
particles,
self.preprocess_image(frame),
self.goodness,
sampling_function=self.sample,
resampling_function=self.save_highest_weight_particle_and_resample,
)
track[i, :] = particles.mean(axis=0)
self.resampled_particles[obj_i][i] = particles
print "Tracked frame %i of %i" % (i + 1, self.num_frames)
self.highest_weight_particles[obj_i] = numpy.array(self.highest_weight_particles[obj_i])
print "Tracking complete."
print
self.tracks[obj_i] = track
def __init__(self, p1, p2, p, w0, ngrain,
filename='temp.txt', dist='g',
iplot=True, idot=False):
path =os.getcwd()
f=open(path+'\\'+filename, 'w')
f.writelines('Designed texture using probability distributions \n')
f.writelines('Given Euler angles are as below \n')
f.writelines('ph1, phi2, PHI = ' + str(p1) + str(p2)+ str(p) )
f.write('\nB '+ str(ngrain))
for i in range(ngrain):
txt = text(p1=p1, p2=p2, p=p, w0=w0, dist=dist)
angle = txt.angle
f.write('\n %9.2f %9.2f %9.2f %9.3f'%(angle[0],angle[1],angle[2], 0.1))
f.close()
if iplot==True:
pf.pf(ftex='temp.txt',idot=idot)
def test_pf(self):
wconf = copy.copy(self.wconf)
wconf.hexTiers = 0
wconf.usersPerCell = 10
wconf.mobileVelocity = 100
world1 = world.World(wconf, self.phy)
world1.associatePathlosses()
world1.calculateSINRs()
rate = 1
avg_rate = np.ones(len(world1.mobiles))
# when avg_rate is near zero for one user, that user 0 should receive all RBs
avg_rate[0] = 1e-20
alloc = pf.pf(world1, world1.cells[0], world1.mobiles, rate, avg_rate)
np.testing.assert_array_equal(alloc, np.zeros([50,10]))
rate = 1e6
pSupplyPC = pf.pf_ba(world1, world1.cells[0], world1.mobiles, rate)
pSupplyDTX = pf.pf_dtx(world1, world1.cells[0], world1.mobiles, rate)
def test_pf(self):
wconf = copy.copy(self.wconf)
wconf.hexTiers = 0
wconf.usersPerCell = 10
wconf.mobileVelocity = 100
world1 = world.World(wconf, self.phy)
world1.associatePathlosses()
world1.calculateSINRs()
rate = 1
avg_rate = np.ones(len(world1.mobiles))
# when avg_rate is near zero for one user, that user 0 should receive all RBs
avg_rate[0] = 1e-20
alloc = pf.pf(world1, world1.cells[0], world1.mobiles, rate, avg_rate)
np.testing.assert_array_equal(alloc, np.zeros([50, 10]))
rate = 1e6
pSupplyPC = pf.pf_ba(world1, world1.cells[0], world1.mobiles, rate)
pSupplyDTX = pf.pf_dtx(world1, world1.cells[0], world1.mobiles, rate)
# -*- coding: utf-8 -*-
"""
Created on Wed Jun 20 14:35:56 2018
@author: jian
"""
from pf import pfaffian, pf
import numpy as np
N = 200
steps = 10
dd = np.zeros(steps)
for i in range(steps):
X = np.random.random((2 * N, 2 * N))
T = X.transpose() - X
dd[i] = abs(pf(T) - pfaffian(T)) / abs(pf(T)) # relative error makes sense
dd.max()
def correlator_dynamics_AABB(self, i, t, j, AABB):
S = self.aux_pfaffian_constructor_t_AABB(i, j, t, AABB)
return pf(S)
def correlator_dynamics(self, i, t, j):
S = self.aux_pfaffian_constructor_t(i, j, t)
return pf(S)
def correlator_equal_time(self, i, j):
S = self.aux_pfaffian_constructor(i, j)
return np.real(pf(S))
def vf(newname, newname_in, newname_out, station_file, final_name):
df2 = pd.read_csv(
newname,
encoding='GBK',
names=['card_id', 'time', 'money', 'line', 'station', 'M1'])
df_in = df2.ix[df2.money == 0.0,
['card_id', 'time', 'money', 'line', 'station', 'M1']]
print(df_in)
df_out = df2.ix[df2.money != 0.0,
['card_id', 'time', 'money', 'line', 'station', 'M1']]
print(df_out)
df_out.columns = [
'card_id', 'time_out', 'money_out', 'line_out', 'station_out', 'M1_out'
]
df_in.columns = [
'card_id', 'time_in', 'money_in', 'line_in', 'station_in', 'M1_in'
]
df_outsh34 = df_out.ix[(df_out.station_out == u'上海火车站') & (
(df_out.line_out == 4) | (df_out.line_out == 3)
), [
'card_id', 'time_out', 'money_out', 'line_out', 'station_out', 'M1_out'
]]
df_insh1 = df_in.ix[
(df_in.station_in == u'上海火车站') & (df_in.line_in == 1),
['card_id', 'time_in', 'money', 'line_in', 'station_in', 'M1_in']]
result34_1 = pd.merge(df_outsh34, df_insh1)
print(df_insh1)
print(df_outsh34)
print(result34_1)
fun = lambda x, y: t1(x) - t1(y)
#result['D']='ColumnD'
#t1(result['time_out'])-t1(result['time_in'])<=60*30
result34_1['duration'] = list(
map(fun, result34_1['time_in'], result34_1['time_out']))
#result[map(lambda x:datetime.datetime(x.year,x.month,x.day,x.hours,x.minutes+30,x.seconds),result['time_in'])>=result['time_out']]
#result.groupby('card_id')
print(result34_1)
result0 = result34_1.ix[
(result34_1.duration <= 60 * 30) & (result34_1.duration > 0), [
'card_id', 'time_out', 'money_out', 'line_out', 'station_out',
'M1_out', 'time_in'
'money_in', 'line_in', 'station_in', 'M1_in', 'duration'
]]
print(result0)
#print(result0.groupby(result0['card_id']).agg({'time_in':['time_in'].min(),'card_id':['card_id']}))
print(result0.groupby(['M1_in', 'card_id'])[['duration']].min())
result1 = result0.groupby(['M1_in', 'card_id'])[['duration']].min()
print("huhu")
#result1['card_id']=result1.index
result1.reset_index('M1_in', inplace=True)
result1.reset_index('card_id', inplace=True)
result2 = pd.merge(result0,
result1,
right_on=['card_id', 'duration'],
left_on=['card_id', 'duration'])
print(
pd.merge(result0,
result1,
right_on=['card_id', 'duration'],
left_on=['card_id', 'duration']))
#print(result0['time_in'].groupby(result0['card_id']).min())
#result1=result0.ix[result0.time_in==result0['duration'].groupby(result0['card_id']).min(),['card_id','money_out','line_out','station_out','time_out','money_in','line_in','station_in','time_in','duration']]
#frames =[result2.ix[:,['card_id','money_out','line_out','station_out','time_out']], df_out0]
#ix[:,['card_id','money_out','line_out','station_out','time_out']
#In [5]: result = pd.concat(frames)
#df_out1=pd.concat(frames, axis=0)
df_out0 = Complement(
result2.ix[:, [
'card_id', 'time_out', 'money_out', 'line_out', 'station_out',
'M1_out'
]], df_out).drop_duplicates()
print(df_out0)
#df_in1=pd.concat([result1['card_id','money_in','line_in','station_in','time_in'],df_in0], axis=0)
df_in0 = Complement(
result2.ix[:, [
'card_id', 'time_in', 'money_in', 'line_in', 'station_in', 'M1_in'
]], df_in).drop_duplicates()
print(df_in0)
###
df_outsh1 = df_out.ix[(df_out.station_out == u'上海火车站') &
(df_out.line_out == 1), [
'card_id', 'time_out', 'money_out', 'line_out',
'station_out', 'M1_out'
]]
df_insh34 = df_in.ix[
(df_in.station_in == u'上海火车站') & ((df_in.line_in == 4) |
(df_in.line_in == 3)),
['card_id', 'time_in', 'money', 'line_in', 'station_in', 'M1_in']]
result1_34 = pd.merge(df_outsh1, df_insh34)
print(df_insh34)
print(df_outsh1)
print(result1_34)
fun = lambda x, y: t1(x) - t1(y)
#result['D']='ColumnD'
#t1(result['time_out'])-t1(result['time_in'])<=60*30
result1_34['duration'] = list(
map(fun, result1_34['time_in'], result1_34['time_out']))
#result[map(lambda x:datetime.datetime(x.year,x.month,x.day,x.hours,x.minutes+30,x.seconds),result['time_in'])>=result['time_out']]
#result.groupby('card_id')
print(result1_34)
result3 = result1_34.ix[(result1_34.duration <= 60 * 30) &
(result1_34.duration > 0), [
'card_id', 'time_out', 'money_out', 'line_out',
'station_out', 'M1_out', 'time_in', 'money_in',
'line_in', 'station_in', 'M1_in', 'duration'
]]
print(result3)
#print(result0.groupby(result0['card_id']).agg({'time_in':['time_in'].min(),'card_id':['card_id']}))
print(result3.groupby(['M1_in', 'card_id'])[['duration']].min())
result4 = result3.groupby(['M1_in', 'card_id'])[['duration']].min()
print("huhu")
#result4['card_id']=result4.index
result4.reset_index('M1_in', inplace=True)
result4.reset_index('card_id', inplace=True)
result5 = pd.merge(result3,
result4,
right_on=['card_id', 'duration'],
left_on=['card_id', 'duration'])
print(
pd.merge(result3,
result4,
right_on=['card_id', 'duration'],
left_on=['card_id', 'duration']))
#print(result0['time_in'].groupby(result0['card_id']).min())
#result1=result0.ix[result0.time_in==result0['duration'].groupby(result0['card_id']).min(),['card_id','money_out','line_out','station_out','time_out','money_in','line_in','station_in','time_in','duration']]
print(result4)
#frames =[result2.ix[:,['card_id','money_out','line_out','station_out','time_out']], df_out0]
#ix[:,['card_id','money_out','line_out','station_out','time_out']
#In [5]: result = pd.concat(frames)
#df_out1=pd.concat(frames, axis=0)
df_out1 = Complement(
result5.ix[:, [
'card_id', 'time_out', 'money_out', 'line_out', 'station_out',
'M1_out'
]], df_out0).drop_duplicates()
print("finaout")
print(df_out1)
#df_in1=pd.concat([result1['card_id','money_in','line_in','station_in','time_in'],df_in0], axis=0)
df_in1 = Complement(
result5.ix[:, [
'card_id', 'time_in', 'money_in', 'line_in', 'station_in', 'M1_in'
]], df_in0).drop_duplicates()
print("finain")
print(df_in1)
df_out1.to_csv(newname_out, encoding='GBK', index=False, header=False)
df_in1.to_csv(newname_in, encoding='GBK', index=False, header=False)
# 调用pf
pf.pf(station_file, newname_in, newname_out, final_name)
