I am trying to parse a dataframe column called 'tags' that contains a list of dicts and as an output create a list of the values of the key var1:

Dataframe column 'tags' example value:

[{'var1': 'blue','var2': 123,'var3': 888},{'var1': 'red','var2': 123,'var3': 888},{'var1': 'green','var2': 123,'var3': 888}]

desired output:

['blue', 'red', 'green']

code:

d = [{f'{k}{i}': v for i, y in enumerate(x, 1) for k, v in y.items() if k == 'var1'} for x in df['tags']]
df = pd.DataFrame(d, index=df.index).sort_index(axis=1)

However this produces the following error:

Type Error: 'float' object is not iterable

I have tried converting both i and v to a string using str(i) and str(v), however I am still getting the same error.

What am I doing wrong?

CodePudding user response：

Error obviously means some missing values instead lists, for avoid it add if-else with empty lists in ouput if missing in tags:

df['new'] = [[] if isinstance(x, float) else [y.get('var1') for y in x] for x in df['tags']]