Pandas and reg ex, decompoising text and numbers into several columns with headings - regex

Question

I have a dataframe with a column containing:
1 Tile 1 up Red 2146 (75) Green 1671 (75)
The numbers 1 can be upto 10
up can be also be down
The 2146 and 1671 can be any digit upto 9999
Whats the best way to break out each of these into separate columns without using split. I was looking at regex but not sure how to handle this (especially the white spaces). I liked the idea of putting the new column names in too and started with
Pixel.str.extract(r'(?P<num1>\d)(?P<text>[Tile])(?P<Tile>\d)')
Thanks for any help

Answer 1

To avoid an overly complicated regex pattern, perhaps you can use str.extractall to get all numbers, and then concat to your current df. For up or down, use str.findall:
df = pd.DataFrame({"title":["1 Tile 1 up Red 2146 (75) Green 1671 (75)",
"10 Tile 10 down Red 9999 (75) Green 9999 (75)"]})
df = pd.concat([df, df["title"].str.extractall(r'(\d+)').unstack().loc[:,0]], axis=1)
df["direction"] = df["title"].str.findall(r"\bup\b|\bdown\b").str[0]
print (df)
#
title 0 1 2 3 4 5 direction
0 1 Tile 1 up Red 2146 (75) Green 1671 (75) 1 1 2146 75 1671 75 up
1 10 Tile 10 down Red 9999 (75) Green 9999 (75) 10 10 9999 75 9999 75 down