How to compare two cols' str elements in dataframes

Question

Currently, I have a dataframe, I want to compare two cols.

col_1 is words, col_2 is a phrase. I want to know, if two words (in col_1) appear in col_2, then change two single words to be a phrase.

here is the dataframe

list1 = [['good', 'hello', 'morning',],['sit', 'good', 'down'],['get', 'who', 'down']]

list2 = [['good morning', 'good afternoon'],['sit down', 'rise up', 'good work'], ['sit here', 'get job', 'get down']]

df_new = pd.DataFrame({'words': list1})

df_new['para'] = list2

I want to get the result is: look like:

list3 = [['good morning', 'hello'],['sit down', 'good'],['get down', 'who']]
list4 = [['good afternoon'],['rise up', 'good work'], ['get job', 'get down']]
df_new['result1'] = list3
df_new['result2'] = list4

• result1: if words appear in phrase, change the two words to one phrase.
• result2: delete the phrase in result1, the remaining phrase in result2.

Any suggestion for how to get :result1,2. I will really grateful for the logic if you could help me! thx for that.

Answer 1

Try to use only list, dtaframe will complicate the task. As for the logic, this is how I would proceed: Using itertools you get every possible pair for your set of word which you compare to your sentences. If it's a match you upload your different lists

import itertools

list1 = [['good', 'hello', 'morning',],['sit', 'good', 'down'],['get', 'who', 'down']]

list2 = [['good morning', 'good afternoon'],['sit down', 'rise up', 'good work'], ['sit here', 'get job', 'get down']]

def possible_pair(list):
    n = len(list)
    possible_pairs = itertools.permutations(list, r=2)
    return possible_pairs

for i,words in enumerate(list1):
    for pair in possible_pair(words):
        sentence = pair[0]+' '+pair[1]
        print(sentence)
        if sentence in list2[i]:
            list2[i].remove(sentence)
            list1[i].append(sentence)
            list1[i].remove(pair[0])
            list1[i].remove(pair[1])

print(list1, list2)

output:

[['hello', 'good morning'], ['good', 'sit down'], ['who', 'get down']] [['good afternoon'], ['rise up', 'good work'], ['sit here', 'get job']]

Once done you can eventually rebuild your df