performance - Multiplication matrix vector with Python 3.x, numpy e multiprocessing -


i'm writing benchmark show differences in time, cost, speed up, efficiency , effective of sequential , parallel algorithm calculate system of linear equations using python 3.5, when calculate different processes, calculation not performed correctly because result of calculation in sequential (and numpy) different done in parallel. noticed half of matrix, each time increasing number of processes, not calculate sets 0 i-th position. doing wrong? mistake , why? code:

import random import sys import time import multiprocessing import math, numpy np numpy import linalg import argparse, sys, os   def creamatrici(n):     = [[random.random() x in range(n)] x in range(n)]     b = [random.random() x in range(n)]     #a^(-1)     = np.linalg.inv(a)     return a,b   def np_moltiplica(a, b):     c = np.dot(a,b)     return c   def moltiplica_seriale(a,b,n):     c = [0 x in range(n)]     in range(0,len(a)):         j in range(0,len(a)):             c[i] += a[i][j] * b[j]     return c  def moltiplica_parallelo(result_queue, worker_index, chunk_boundary, a,b):     n = len(a)     result = []     m in range(n):         output = moltiplica_seriale(                 a[chunk_boundary[0]:chunk_boundary[1],:],b,n)         result.append(output)      result_queue.put((worker_index, result))  ###########################################################################  def divisione_lavoro(nworkers, a,b):     def calcola_partizionamento_lavoro(l, n):         c = int(math.ceil(len(l)/float(n)))         return c      sizepartizionamento = calcola_partizionamento_lavoro(a, nworkers)     limitimatricidivise = [(i, i+sizepartizionamento) in range(0, (len(a)), sizepartizionamento)]     if nworkers != len(limitimatricidivise):         return none      result_queue = multiprocessing.queue()     processifigli = []     worker_index in range(nworkers):         processifigli.append(                                          multiprocessing.process(target=moltiplica_parallelo,                               args=(                                                                                                                    result_queue,                                    worker_index,                                    limitimatricidivise[worker_index],                                    a,                                     b)                                       )                            )      c in processifigli:         c.start()      results = [0 n in range(len(a))]      _ in range(nworkers):         worker_index, result_chunk = result_queue.get(block=true)         chunk_boundary = limitimatricidivise[worker_index]         results = result_chunk[0]       c in processifigli:         c.join()     return results       ###########################################################################      def main(argv=none):     if argv none:         argv = sys.argv      n = int(argv[1])       a,b = creamatrici(n)         #########################seriale##########################      start = time.time()     c = moltiplica_seriale(a,b,n)     end = time.time()       duration = (end-start) % 60     print ('risultato = ', c)     print ('\ntempo seriale: %.5f s' % duration)      #########################numpy############################      start = time.time()     c = np_moltiplica(a,b)     end = time.time()      duration = (end-start) % 60      print ('\n\nrisultato = ', c)     print ('\ntempo usando il modulo numpy: %.5f s' %duration)     print ('\n\n')      ######################multiprocessing#######################      durations = []     n_children = [1, 2, 4, 8]     n in n_children:         start = time.time()         result = divisione_lavoro(n, a, b)         if result none:             continue         duration = (time.time() - start) %60         if n == 1:             print ("risultato calcolato con %s processo figlio: %s" % (n, result))         else:             print ("risultato calcolato con %s processi figli: %s" % (n, result))         print ("tempo: %.2f s" % duration)         print ('\n\n')         durations.append(duration)      costo = [n*d n,d in zip(n_children,durations)]     speedup = [durations[0]/d d in durations]     efficienza = [s/n s,n in zip(speedup, n_children)]     efficacia = [s/c s,c in zip(speedup,costo)]     n, costo, speed, efficienza, efficacia in zip(n_children,costo,speedup,efficienza,efficacia):         if n == 1:             print ("costo con %s-processo figlio: %.2f" % (n,costo))             print ("speedup con %s-processo figlio: %.2f" % (n, speed))             print ("efficienza con %s-processo figlio: %.2f" % (n, efficienza))             print ("efficacia con %s-processo figlio: %.2f\n" % (n, efficacia))         else:             print ("costo con %s-processi figli: %.2f" % (n,costo))             print ("speedup con %s-processi figli: %.2f" % (n, speed))             print ("efficienza con %s-processi figli: %.2f" % (n, efficienza))             print ("efficacia con %s-processi figli: %.2f\n" % (n, efficacia))  if __name__ == "__main__":     sys.exit(main())


Comments

Post a Comment

Popular posts from this blog

authentication - Mongodb revoke acccess to connect test database -

i have ec2 instance mongodb 3.2.0 installed. have user admin access. can connect user , perform operation remotely. 1 can connect database test remote if knows ip mongo 11.11.11.11. although not able make operations, want restrict access user credentials can connect server. first need create users , grant roles users, example: use admin db.createuser( { user: "myuseradmin", pwd: "abc123", roles: [ { role: "useradminanydatabase", db: "admin" } ] } ) then need start server --auth parameter mongod --auth more information here
Read more

r - Update two sets of radiobuttons reactively - shiny -

i have read ( how make choices in radiobuttons reactive in shiny? ) shows me how update radiobutton s in reactive way. however, when try , update 2 sets of buttons same data, 1 set renders. example: server: # create example data wafer <- rep(c(1:3), each=3) length <- c(1,1,2,1,1,1,3,5,1) width <- c(3,1,6,1,1,1,1,1,6) dd <- data.frame(wafer, length, width) shinyserver(function(input, output, session){ # create reactive dataframe store data values <- reactivevalues() values$df <- data.frame() # lengths , widths of wafer user input <- eventreactive(input$do, { subset(dd, wafer %in% input$wafer, select = length:width) }) # update reactive data frame lengths , widths have been selected user input observe({ if(!is.null(a())) { values$df <- rbind(isolate(values$df), a()) } }) output$wl <- renderdatatable({ a() }) # update radio buttons unique length , widths stored in values$df # ever "observe" put first in co...
Read more

ios - Realm over CoreData should I use NSFetchedResultController or a Dictionary? -

i'm working on app using core data , nsfetchedresultcontroller many reasons switch use realm. saw there "realm version" of nsfetchedresultcontroller on github wouldn't compatible current code using core data. the view controller displaying list of people same school, address book. list sublist of people studied in same city. so thinking make 1 unique request database retrieve list of people , filter locally list within 1 dictionary per school [string: anyobject] string section name within tableview , anyobject array of people. first solution, make nsfetchedresultcontroller each school predicate filtering location. delete actions etc.. handled delegates -> not compatible realm create dictionaries , update them each actions... -> works realm it's annoying code. any better solution? edit: i need clarify request: i'd write class inherit uitableviewcontroller. this class has list of people sorted in alphabetical order the tableview has ...
Read more