How to append empty row/column and how to use .empty() method properly?

user_11889874 0
Viewed 37

I'm writing a function which takes item from a list (such as item_x in [item_1, item_2, item_3, ..., item_n]) and returns dataframe_i (can be empty).

I want to concatenate the returned dataframes into a single final_result_df, which also has the same length as the list (even all returned dataframes are empty). I need to check if the final_result_df is empty.

Here are my codes:

def get_content_df(item):
    content = do_something()
    if content is not None:
        result_df = pd.DataFrame([content])
    else:
        result_df = pd.DataFrame([""])  # This line is my problem
    return result_df

my_list = [item_1, item_2, item_3, ..., item_n]

final_result_df = pd.DataFrame()
for item in my_list:
    result_df_i = get_conetent_df(item)
    final_result_df = pd.concat([final_result_df, result_df_i], ignore_index=True, axis=0)

print(final_result_df.empty)

The problem happens when all result_df_i fall into the condition that content is None is True.

  1. If I use result_df = pd.DataFrame([""]), the final_result_df has the shape I want, but the .empty method returns the wrong value.

  2. If I use result_df = pd.DataFrame(), the final_result_df does not have the shape I want, but the .empty method returns the correct value.

I believe the first condition must have some quick fix, but I'm not sure how to achieve that. I know this might not be how pandas was supposed to be used, but I haven't came up with a better idea. Any advice on other methods is also wellcomed.

pandas python
1 Answers

try:

def get_content_df(item):
    if item is nan:
        item = "" 
    content = item
    if content is not None:
        result_df = pd.DataFrame([content])
    else:
        result_df = pd.DataFrame([""])
    return result_df

I this is ok !, no?

my_list = ["", ""]
my_list2 = [22, "dksjds"]


for j in my_list2:
    #print(i, get_content_df(i))
    res = get_content_df(j)
    #print(res)
    final_result_df3 = pd.concat([final_result_df3, res], ignore_index=True, axis=0)

even if the list is a series of nan, or ""s - it gets the result.

i.e. if list = [None] * 10, say it works...

hope this helps ...

user_101453980
Related