从DF列将值插入字典-Pandas（Python）

Question

I have a list of values that I am trying to match with a column of a pandas df and then would like to create a dictionary that will have list values as keys and then dictionary values from a different column from the data frame.

This is how I have my list:

sample_list = [101,105,112]

My Data Frame:

sample_df = pd.DataFrame([[101, "NJ"], [105, "CA"],[111, "MO"], [101, "NJ"], [112, "NB"], [101, "NJ"], [105, "CA"]], \
                         columns=["Col1", "Col2"])

looks like this,

    Col1    Col2
0   101     NJ
1   105     CA
2   111     MO
3   101     NJ
4   112     NB
5   101     NJ
6   105     CA

Now, I am trying to iterate list values (which are keys of my new_dict )and match them with Col1 and if they match I would like to extract Col2 values as my dictionary values. This is how I have my code so far,

new_dict = {}
for value in sample_list:
    for i in sample_df['Col1']:
        if value == i:
            new_dict[value] = [i for i in sample_df['Col2']]

However, my new_dict looks like this,

{101: ['NJ', 'CA', 'MO', 'NJ', 'NB', 'NJ', 'CA'],
 105: ['NJ', 'CA', 'MO', 'NJ', 'NB', 'NJ', 'CA'],
 112: ['NJ', 'CA', 'MO', 'NJ', 'NB', 'NJ', 'CA']}

I need my output like this,

{101: ['NJ'],
 105: ['CA'],
 112: ['NB']}

How can I get to my desired output? Any help would be nice.

Answer 1

这样做：

new_dict = {i: [sample_df[sample_df['Col1']==i]['Col2'].values[0]] for i in sample_list}

Answer 2

Alt 1

If you insist here is another solution that should be efficient by using isin() to create a mask used to filter away not desired rows.

m = sample_df['Col1'].isin(sample_list)
sample_df[m].drop_duplicates().groupby('Col1')['Col2'].apply(list).to_dict()

Returns: {101: ['NJ'], 105: ['CA'], 112: ['NB']}

note: if there are more non-unique combos they will be in the list too . Use: {k:[v] for k,v in sample_df[m].groupby('Col1')['Col2'].first().items()} if you only want the first.

---

Alt 2

If you are going for list items but not all why not just the values?

m = sample_df['Col1'].isin(sample_list)
sample_df[m].set_index('Col1')['Col2'].to_dict()

Returns: {101: 'NJ', 105: 'CA', 112: 'NB'}

---

Alt 3

or if you want all the items:

m = sample_df['Col1'].isin(sample_list)
sample_df[m].groupby('Col1')['Col2'].apply(list).to_dict()

Returns: {101: ['NJ', 'NJ', 'NJ'], 105: ['CA', 'CA'], 112: ['NB']}