Dynamically append dataframes in pandas

Question

I want to load files from a list, calculate mean, median and standard deviation for each row of each file and then create a dataframe listing all the newly calculated fields.

I have the following code:

    #list files to load
    file_names = ["file_1", "file_2", ...]

    #empty df
    data = pd.DataFrame()

    #for loop
    for filename in file_names:
        df = pd.read_csv(filename, index_col=False, header=0)
        mean = df.mean(axis = 1)
        median = df.median(axis = 1)
        std = df.std(axis = 1)
        df = pd.concat([mean, median, std], axis = 1, ignore_index = 1)
    data = pd.concat(df, axis=1)

I'm getting an error:

TypeError: first argument must be an iterable of pandas objects, you passed an object of type "DataFrame"

Individual dfs that are being created in the for loop look exactly how I want it but I can't concatenate them all together.

Answer 1

As it is you're overwriting df every time through the loop.
Instead collect the DataFrames in a list, then concatenate that list together.

df_list = []

#for loop
for filename in file_names:
    df = pd.read_csv(filename, index_col=False, header=0)
    mean = df.mean(axis = 1)
    median = df.median(axis = 1)
    std = df.std(axis = 1)
    df = pd.concat([mean, median, std], axis = 1, ignore_index = 1)
    df_list.append(df)

data = pd.concat(df_list, axis=1)