Every element of the array "data" have to be changed as follows:
For example, 4 should be seen in names_A and data_A. The names_A for 4 is 'David'. Now 'David' should be seen in names_B and data_B. The data_B for 'David' is 30. So, the element 4 must be changed by 30; and so on.
import numpy as np
names_A = ['David', 'Mark', 'Brian', 'Michael']
data_A = [4,3,1,2]
names_B = ['Mark', 'David', 'Michael', 'Brian']
data_B = [51,30,11,29]
data = np.array([[4,4,3,3,2,2,1,1,3,3],
[4,3,3,3,2,2,3,1,3,1],
[4,2,3,3,2,2,4,1,4,3]])
How is the easiest and simplest way of doing it?
I tried it as follows:
dats = data.ravel()
results = []
for d in dats:
nam_A = names_A[data_A == int(d)]
dat_B = data_B[names_B == nam_A]
results.append(dat_B)
print np.array(results).reshape(data.shape)
[[51 51 51 51 51 51 51 51 51 51]
[51 51 51 51 51 51 51 51 51 51]
[51 51 51 51 51 51 51 51 51 51]]
But, it's giving wrong results. How would you do it?
Use dictionaries to create a mapping.
names_A = ['David', 'Mark', 'Brian', 'Michael']
data_A = [4,3,1,2]
names_B = ['Mark', 'David', 'Michael', 'Brian']
data_B = [51,30,11,29]
lookup_a = dict(zip(names_A, data_A))
lookup_b = dict(zip(names_B, data_B))
mapping = {value_a: lookup_b[key_a] for key_a, value_a in lookup_a.items()}
Now the keys in mapping
will be the numbers from data_A
with the corresponding values from data_B
.
I never worked with numpy
but it looks like an easy task to do the replacement now.
Just to give an example with a simple list:
data = [4, 4, 3, 3, 2, 2, 1, 1, 3, 3]
data = [mapping[value] for value in data]
data
now is [30, 30, 51, 51, 11, 11, 29, 29, 51, 51]
.
Edited after installing numpy
If you created the mapping dictionary you can do the following:
data = np.array([[4, 4, 3, 3, 2, 2, 1, 1, 3, 3],
[4, 3, 3, 3, 2, 2, 3, 1, 3, 1],
[4, 2, 3, 3, 2, 2, 4, 1, 4, 3]])
for row in data:
for index, value in enumerate(row):
row[index] = mapping[value]
data
is now:
[[32 30 51 51 11 11 29 29 51 51]
[30 51 51 51 11 11 51 29 51 29]
[30 11 51 51 11 11 30 29 30 51]]
As I never worked with numpy before there might be easier (or more pythonic) solutions, but at least this does what it should do.
来源:https://stackoverflow.com/questions/30704520/elements-arrangement-in-python