Python字典-将数组附加到特定键的字典中。

Question

This is about appending arrays to a dictionary with a specific key.

Aim is to assign a number of arrays for a specific key in a dictionary:

{1: [1,2,3,4], [6,3,2,3], 2: None, 3: None}

There's a dictionary

my_dict = dict.fromkeys([i for i in range(1,4)]
my_dict
{1: None, 2: None, 3: None}

I have a number of arrays:

array1 = [1,2,3,4]  
array2 = [6,3,2,3]
array3 = [5,7,11,15]

I wish to associate array1 with key = 1

my_dict[1] = array1
#the content of the my_dict is:
my_dict
{1: [1, 2, 3, 4], 2: None, 3: None}

Then I wish to add array2 associated with key 1. BUT if I do this

my_dict[1].append(array2); 
#I get
my_dict
{1: [1, 2, 3, 4, [6,3,2,3]], 2: None, 3: None}

In other words I would like to know how to append for a specific key such that I get the following

{1: [1, 2, 3, 4], [6, 3, 2, 3], 2: None, 3: None}

NOT

{1: [1, 2, 3, 4, [6,3,2,3]], 2: None, 3: None}

In sum , I would like to append a number of arrays to a dictionary for a specific key such that I get:

{1: [1,2,3,4], [6,3,2,3], 2: None, 3: None}

Thank you, Anthony of Sydney.

Answer 1

Your syntax is incorrect. What you likely need is a defaultdict of lists.

Then just use list.append each time you wish to add to a list of lists.

from collections import defaultdict

array1 = [1,2,3,4]  
array2 = [6,3,2,3]
array3 = [5,7,11,15]

my_dict = defaultdict(list)
my_dict[1].append(array1)
my_dict[1].append(array2)

defaultdict(list, {1: [[1, 2, 3, 4],
                       [6, 3, 2, 3]]})

---

If, on the other hand, you want a single list and just wish to add elements, use list.extend :

my_dict = defaultdict(list)
my_dict[1].extend(array1)
my_dict[1].extend(array2)

# defaultdict(list, {1: [1, 2, 3, 4, 6, 3, 2, 3]})

Answer 2

I do not think you can achieve it.

{1: [1,2,3,4], [6,3,2,3], 2: None, 3: None}

but you can get:

{1: [[1,2,3,4], [6,3,2,3]], 2: [], 3: []}

my_dict = {i:[]for i in range(1,4)}
my_dict[1].append(array1)
my_dict[1].append(array2)

Answer 3

I'm guessing that what you want is {1: [[1, 2, 3, 4], [6, 3, 2, 3]], 2: None, 3: None} . Then you just have to do: my_dict[1] = [array1, array2] . Otherwise if what you want is {1: [1, 2, 3, 4, 6, 3, 2, 3], 2: None, 3: None} you have to do: my_dict[1] += array2 (once you have already done my_dict[1] = array1 ).

Answer 4

This is an expansion of the first answerer and my comment to the first answerer.

If I have the following arrays:

array1 = [1,2,3,4]
array2 = [2,6,27,1]
array3 = [6,1,2,3]
array4 = [7,1,2,6]
array5 = [8,2,1,6]
array6 = [3,4,1,1]

My dictionary is the following:

from collections import defaultdict
my_dict = defaultdict(list)

Now I will append the following arrays for different keys:

my_dict[1].append(array1)
my_dict[1].append(array2)
my_dict[1].append(array3)
my_dict[2].append(array4)
my_dict[3].append(array5)
my_dict[3].append(array6)

my_dict's contents

{1:[[1,2,3,4],[2,6,27,1],[6,1,2,3]], 2:[[7,1,2,6]], 3:[[8,2,1,6],[3,4,1,1]] }

To get the array associated with key = 1

foo = my_dict[1]
foo
[[1,2,3,4],[2,6,27,1],[6,1,2,3]]

Now to get the 3rd column of data:

import numpy as np #Need to use it to convert to array and float later on
foo = np.array(foo);# Have to convert it to an array
doo = foo[:,2]
doo
[3,27,2]

Now there may be situations where your array may be a mix of string and float or integer such that all elements in the array are string.

foo 
[['1','2','3','4'],['2','6','27','1'],['6','1','2','3']]
foo = np.float32(foo)
foo
[[1.,2.,3.,4.],[2.,6.,27.,1.],[6.,1.,2.,3.]]

Conclusion: We can append arrays to a dictionary for a specific key using the defaultdict. Also subsets of the selected matrix can be extracted via normal column selection method. If the array associated with the key contains string elements, we can convert the data to numpy.float32, say or wherever we need to do computation with numbers.

Thank you, Anthony of Sydney