Python updating one key also updating other key in dictionary-CodePudding

val = {
    '1': {
      'amt': 150.0,
      'qty': 10
    }
  }

ops = {
  'add': val,
  'remove': val
}

new = {'2': {'amt': 0.0, 'qty': 5}}

ops["add"].update(new)

print(ops)

when you run this, the output is,

{'add': {'1': {'amt': 150.0, 'qty': 10}, '2': {'amt': 0.0, 'qty': 5}}, 'remove': {'1': {'amt': 150.0, 'qty': 10}, '2': {'amt': 0.0, 'qty': 5}}}

So when I update the key 'add' its also updating 'remove' with value 'new'. I am guessing its due to the 'shallow copy' thing. But how to handle this in better way ?

CodePudding user response：

Use deepcopy to copy nested mutable structures:

from copy import deepcopy

ops = {'add': deepcopy(val),
       'remove': deepcopy(val)}

ops['add'].update(new)

Output:

In [3]: ops
Out[3]:
{'add': {'1': {'amt': 150.0, 'qty': 10}, '2': {'amt': 0.0, 'qty': 5}},
 'remove': {'1': {'amt': 150.0, 'qty': 10}}}

In [4]: val
Out[4]: {'1': {'amt': 150.0, 'qty': 10}}

CodePudding user response：

Your issue is that val is the same object for both elements of ops. Try this:

val = {
    '1': {
      'amt': 150.0,
      'qty': 10
    }
  }

ops = {
  'add': val.copy(),
  'remove': val.copy()
}

new = {'2': {'amt': 0.0, 'qty': 5}}

ops["add"].update(new)

print(ops)

CodePudding user response：

Ah yes I got it, sorry for mistaken your question. I think you have to copy it and write in the new address (in C/C what you did was using the same pointer (pointing the same memory address))

you can try this

val = {
    '1': {
      'amt': 150.0,
      'qty': 10
    }
  }

ops = {
  'add': val.copy(),
  'remove': val.copy()
}

new = {'2': {'amt': 0.0, 'qty': 5}}

ops["add"].update(new)

by using copy() you're implicitly write the same dict in different memory address