Aggregation conditions in field1, if the value in field2

Question

We introduce some examples of data:

curl -XPUT localhost:9200/dt/art/1 -d '{ "basket1": ["banana","apple", "peach"], 
                                         "basket2": ["banana", "orange"]}'
curl -XPUT localhost:9200/dt/art/2 -d '{ "basket1":["orange", "apple"],
                                         "basket2":[]}'
curl -XPUT localhost:9200/dt/art/3 -d '{ "basket1": ["apple"],
                                         "basket2": ["apple", "banana"]}'
curl -XPUT localhost:9200/dt/art/4 -d '{ "basket1:["banana"],
                                         "basket2": ["banana"]}'

I need to find the article number where the “banana” was found in basket1 but not in basket2. and repeat this for all values ​​in basket1. For example, I would expect here:

banana : 0
apple: 2
peach: 1
orange: 1

So far, the only solution I have found is to sub-aggregate:

body = {
        "query": {
            "match_all": {}},
        "aggs": {
            "basket1_count": {
                "terms":{
                    "field": basket1},
                "aggs": {
                    "basket2_count": {
                        "terms":{
                            "field": basket2,
                            "size": 10000}}}}}}

Then I would stay in python. (the difference between the number of documents where the value in field 1 is found and the number of documents where the value in field 2 is found)

Is there a better way to do this?