I have the list:
[['abc', 1, 2, 3], ['bfg', 4, 5, 6], ['abc', 7, 8, 9], ['bfg', 10, 11, 12]] And I make it into a pandas DataFrame, which returns (after adding a column with the color lst[4] = 'blue'):
0 1 2 3 4 0 abc 1 2 3 blue 1 bfg 4 5 6 blue 2 abc 7 8 9 blue 3 bfg 10 11 12 blue Is there anyway to make it return this instead:
0 1 2 3 4 0 abc 1 2 3 blue 1 bfg 4 5 6 blue 2 abc 7 8 9 red 3 bfg 10 11 12 red Solution with DataFrame.from_records:
lst = [['abc', 1, 2, 3], ['bfg', 4, 5, 6], ['abc', 7, 8, 9], ['bfg', 10, 11, 12]] df = pd.DataFrame.from_records(lst) print (df) 0 1 2 3 0 abc 1 2 3 1 bfg 4 5 6 2 abc 7 8 9 3 bfg 10 11 12 Add values by loc:
l = len(df.index) // 2 df.loc[:l - 1, 4] = 'blue' df.loc[l:, 4] = 'red' print (df) 0 1 2 3 4 0 abc 1 2 3 blue 1 bfg 4 5 6 blue 2 abc 7 8 9 red 3 bfg 10 11 12 red More interesting is if there is odd length of df - floor division //:
lst = [['abc', 1, 2, 3], ['bfg', 4, 5, 6], ['abc', 7, 8, 9], ['bfg', 10, 11, 12], ['bfg', 3, 4, 5]] df = pd.DataFrame.from_records(lst) print (df) 0 1 2 3 0 abc 1 2 3 1 bfg 4 5 6 2 abc 7 8 9 3 bfg 10 11 12 4 bfg 3 4 5 l = len(df.index) // 2 df.loc[:l, 4] = 'blue' df.loc[l:, 4] = 'red' print (df) 0 1 2 3 4 0 abc 1 2 3 blue 1 bfg 4 5 6 blue 2 abc 7 8 9 red 3 bfg 10 11 12 red 4 bfg 3 4 5 red or normal division /:
l = len(df.index) / 2 df.loc[:l, 4] = 'blue' df.loc[l:, 4] = 'red' print (df) 0 1 2 3 4 0 abc 1 2 3 blue 1 bfg 4 5 6 blue 2 abc 7 8 9 blue 3 bfg 10 11 12 red 4 bfg 3 4 5 red Simplest is:
lst.loc[len(lst)/2:, 4] = 'red' 