如何在循环内将数据框追加到现有数据框

Question

I made a simple DataFrame named middle_dataframe in python which looks like this and only has one row of data: display of the existing dataframe And I want to append a new dataframe generated each time in a loop to this existing dataframe. This is my program:

    k = 2
    for k in range(2, 32021):
        header = whole_seq_data[k]
        if header.startswith('>'):
            id_name = get_ucsc_ids(header)
            (chromosome, start_p, end_p) = get_chr_coordinates_from_string(header)
        if whole_seq_data[k + 1].startswith('[ATGC]'):
            seq = whole_seq_data[k + 1]
        df_temp = pd.DataFrame(
            {
                "ucsc_id":[id_name],
                "chromosome":[chromosome],
                "start_position":[start_p],
                "end_position":[end_p],
                "whole_sequence":[seq]
            }
        )
        middle_dataframe.append(df_temp)
        k = k + 2

My iterations in the for loop seems to be fine and I checked the variables that stored the correct value after using regular expression. But the middle_dataframe doesn`t have any changes. And I can not figure out why.

Answer 1

The DataFrame.append method returns the result of the append, rather than appending in-place ( link to the official docs on append ). The fix should be to replace this line:

        middle_dataframe.append(df_temp)

with this:

    middle_dataframe = middle_dataframe.append(df_temp)

Depending on how that works with your data, you might need also to pass in the parameter ignore_index=True .

The docs warn that appending one row at a time to a DataFrame can be more computationally intensive than building a python list and converting it into a DataFrame all at once. That's something to look into if your current approach ends up too slow for your purposes.