How to remove duplicate number in the sub list, And retain maximum duplicate number content in sub list

Revit
1

"Hi Dynamo Community,

Is it possible to identify duplicate numbers within sublists and retain only the maximum duplicate number content in each sublist?

For example, in the first image, the actual list is as shown below:

sample 1
sample 1881Ă—871 56.7 KB

And in the second image, I want the content to be as shown below:

sample
sample767Ă—871 54.2 KB

Thank you."

2

@sjafarali75

check out this topic, i think you can modify the issue to your aims…

3

Hi @Draxl_Andreas,

Thank you for your response. However, the issue I’m encountering is within the sublists, not the main list. Additionally, I’m not very familiar with Python. Could you please assist me with this?

4

I tried using ChatGPT for this issue, and it provided some code. However, the code it generated produced the same result as the original list.

sample 0
sample 01457Ă—936 78.8 KB

5

ChatGPT is likely going to struggle with bespoke functionality around specific use cases and would require some very specific language to describe what it is exactly you’re looking for. You can tell from looking at the code that it didn’t understand the request (it’s looking for the maximum value duplicate) and still gave bad code (it removes the item from the sublist only to add it back on the very next line).

What ChatGPT could be helpful with, and is the general exercise I’d suggest for something like this, is to break down and define all the individual parts of your condition first. See if you can:

  1. Identify duplicate values in sublists (either by index or boolean mask).
  2. Identify max values from each duplicate list (either by index or boolean mask).
  3. Identify first instances of each duplicate value (either by index or boolean mask).
  4. Cross-reference each condition for final list of duplicate values to keep and/or which to get rid of.

Python would make this much easier, but if you can work through most of the items above, we can probably help you write something with nodes that can get the trick done.

1 Like
6

7

Watch out, people. AI is coming for your job!

/s

1 Like
8 
import copy

def remove_dupes(list_of_lists):
    seen = set()
    working_list = copy.deepcopy(list_of_lists)
    for i, sub_list in enumerate(list_of_lists):
        index = []
        for j, val in enumerate(sub_list):
            if val in seen:
                index.append(j)
            else:
                seen.add(val)
        for idx in reversed(index):
            working_list[i].pop(idx)
    return list(filter(None, working_list))

A few tips here

  • Python will modify the original object so make a deepcopy to ensure original list and sub-lists are maintained.
  • The pop (extracting a value) method alters a list length so if we work backwards the indexing is maintained. This may have some issues if your sublists have duplicates so we check all the values L → R and remove them in reverse.
  • The filter removes any empty lists - however in Python3 this is a generator so running list on it will expand the result (Dynamo also does this on processing OUT) :robot:
9

Hi @Mike.Buttery,

Thank you for your response.

The code worked, but it removed duplicate items from each sublist, making them unique sublists. Allow me to clarify what I’m looking for.

In Picture 1, you’ll see my original list.

sample 0
sample 0382Ă—773 25.2 KB

What I want is to join the sublists if there are duplicate numbers present, as shown in Picture 2 or 3.

sample
sample481Ă—703 16 KB

sample 1
sample 1375Ă—598 11 KB

10

This is a exercise for List.GroupByKey. The list is the list as you have it now.

The hard part is creating a key that is a comparable element for each group. A String From Array node with a List.Map method might work there.

11

Can you state you logic clearly - how does it handle these numbers indicated?

image
image554Ă—897 58.6 KB

12

Hi @Mike.Buttery ,
My apologies for the confusion. I’ve now corrected the image. Could you please take a look and assist accordingly?

new
new546Ă—758 36.4 KB

13

So as I understand it the logic is
For a list of lists

  • Processing each list from left to right
  • If the maximum number in a list is duplicated in other lists, append the other lists
import copy

def group_by_max(list_of_lists):
    working_list = copy.deepcopy(list_of_lists)
    result = []

    while working_list:
        sub_list = working_list.pop(0)
        collector = [sub_list]

        for i, next_list in enumerate(working_list):
            if max(sub_list) in next_list:
                collector.append(working_list.pop(i))

        if len(collector) > 1:
            result.append(collector)
        else:
            result.append(sub_list)

    return result
14

Hi @Mike.Buttery,

Sorry for the late reply. Yes, this code provided the perfect solution.
Thanks!

1 Like