为什么我们使用Hadoop MapReduce进行数据处理?为什么不在本地机器上做?

Question

我很困惑，我试图把概率当作一百万个随机数。 我尝试了两种方式，在谷歌dataProc中使用MapReduce，并在spyder上运行python脚本来做同样的事情。 但本地机器的速度越快。 那我们为什么使用Mapreduce呢？ 我使用以下代码。

#!/usr/bin/env python3
import timeit
start = timeit.default_timer()
from collections import Counter
import numpy as np
import matplotlib.pyplot as plt
#Random Number Generating
x = np.random.randint(low=1, high=100, size=1000000)
counts = Counter(x)
total = sum(counts.values())
d1 = {k:v/total for k,v in counts.items()}
grad = d1.keys()
prob = d1.values()
#print(str(grad))
#print(str(prob))
#bins = 20
plt.hist(prob,bins=20, normed=1, facecolor='blue', alpha=0.5)
#plt.plot(bins, hist, 'r--')
plt.xlabel('Probability')
plt.ylabel('Number Of Students')
plt.title('Histogram of Students Grade')
plt.subplots_adjust(left=0.15)
plt.show()
stop = timeit.default_timer()
print('Time: ', stop - start)

#!/usr/bin/env python3
"""mapper.py"""
import sys
# Get input lines from stdin
for line in sys.stdin:
# Remove spaces from beginning and end of the line
#line = line.strip()
# Split it into tokens
#tokens = line.split()
#Get probability_mass values
for probability_mass in line:
print("Nonet{}".format(probability_mass))
#print(str(probability_mass)+ 't1')
#print('%st%s' % (probability_mass, None))

#!/usr/bin/env python3
"""reducer.py"""
import sys
from collections import defaultdict
counts = defaultdict(float)
# Get input from stdin
for line in sys.stdin:
#Remove spaces from beginning and end of the line
#line = line.strip()
# skip empty lines
if not line:
continue  
# parse the input from mapper.py
k,v = line.split('t', 1)
counts[v] += 1
total = (float(sum(counts.values())))
#total = sum(counts.values())
probability_mass = {k:v/total for k,v in counts.items()}
print(probability_mass)

Answer 1

Hadoop用于存储和处理大数据。在Hadoop中，数据存储在作为集群运行的廉价商品服务器上。它是一个分布式文件系统，允许并发处理和容错。Hadoop MapReduce编程模型用于更快地存储和从其节点检索数据。

Google Dataproc 是 Apache Hadoop on Cloud。当体积很大时，单台机器不足以处理映射/减少。 100万是小体积。