![](/img/trans.png)
[英]How can I multiply matrices of matrices elementwise using numpy (and not for loops)?
[英]How can I multiply matrices in python using MapReduce?
假设我有两个如下所示的 2X2 矩阵。
A,0,0,1
A,0,1,0
A,1,0,0
A,1,1,1
B,0,0,2
B,0,1,3
B,1,0,4
B,1,1,5
例如B,1,0,4
表示矩阵 B,第 1 行,第 0 列,值 4。
我如何使用 MapReduce 方法计算两个指标的乘积,以获得此 output:
0,0,2
0,1,3
1,0,4
1,1,5
我制作了一个cache.txt
文件来输入矩阵的维度。 (row_a, col_b, col_a)
input.txt
包含矩阵。
代码如下。
cache_info = open("cache.txt").readlines()[0].split(",")
row_a, col_b, col_a = map(int, cache_info)
mapperOutput = open("mapperOutput.txt", "w")
for line in open("input.txt"):
matrix_index, row, col, value = line.rstrip().split(",")
if matrix_index == "A":
for i in range(0, col_b):
key = row + "," + str(i)
mapperOutput.write("%s\t%s\t%s" % (key, col, value) + "\n")
else:
for j in range(0, row_a):
key = str(j) + "," + col
mapperOutput.write("%s\t%s\t%s" % (key, row, value) + "\n")
mapperOutput.close()
numbers1 = list()
for line in open("mapperOutput.txt"):
curr_index, index, value = line.rstrip().split("\t")
index, value = map(int, [index, value])
numbers1.append((curr_index, index, value))
numbers2 = numbers1
initValue1 = list()
initValue2 = list()
for i in numbers1:
checker = 0
for j in numbers2:
if i == j:
if checker == 0:
checker += 1
continue
if i[0] == j[0]:
if i[1] == j[1]:
initValue1.append([i[0],str(i[1]),i[2]*j[2]])
initValue2 = initValue1
myOut = dict()
counter = 0
for i in initValue1:
if counter > (row_a*col_b*col_a):
break
if i[0] in myOut.keys():
counter += 1
continue
counter += 1
myOut[i[0]] = i[2]
inercounter = 0
for j in initValue2:
if i[0] == j[0]:
if i[1] != j[1]:
inercounter += 1
if inercounter > col_b - 1:
continue
myOut[i[0]] += j[2]
for key,value in myOut.items():
print(key,value)
output 将是:
0,0 2
0,1 3
1,0 4
1,1 5
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.