Sort dataframe based on first digit of a column

Question

I have a dataframe like this:

        Produtos    Estoque total     Valor Total de estoque
0            70        10000                  7180
1            70      2800000               2011550
2            70       125000                 89800
3            71       540000                530980
4            71        89000                 79280
5            84       205000                572770
...         ...           ...                    ...
14988   1003254        46000               1329400
14989   1003273     30570000               5502600
14990   1003274     62000000               3720000
14991   1003275    200000000               3840000
14992   1003276       710000               2108700
14993   1003279      6750000                715330

I am trying to sort the 'Produtos' column taking in consideration the first digit first (and if equality, the second, and so on), like:

100
1001
1002
10003
10004
100000
200
2001
20002

I found that I should use this command:

line.sort(key=lambda line: int(line.split()[0]))

But I am having difficult to use in a correct way.

Answer 1

As usual, answer from @jezrael is detailed and superb. But I found another option available and I think it's worth sharing:

For Pandas version >= 1.1:

We can just use key arg in sort_values:

df.sort_values(by = ['Estoque'], key = lambda x:x.astype(str).str[0])

(Sorry I couldn't get the columns lined up with pd.read_clipboard())

The result

Answer 2

First get values of first values by indexing by str, get positions by argsort and last reorder by iloc:

df = df.iloc[df['Produtos'].astype(str).str[0].argsort()]
print (df)
       Produtos  Estoque total  Valor Total de estoque
14988   1003254          46000                 1329400
14989   1003273       30570000                 5502600
14990   1003274       62000000                 3720000
14991   1003275      200000000                 3840000
14992   1003276         710000                 2108700
14993   1003279        6750000                  715330
0            70          10000                    7180
1            70        2800000                 2011550
2            70         125000                   89800
3            71         540000                  530980
4            71          89000                   79280
5            84         205000                  572770

EDIT: Sorting by first value and also length is possible this trick - select by indices by helper DataFrame by first value with len:

print (df)
          Produtos  Estoque total  Valor Total de estoque
0               70          10000                  7180.0
1               70        2800000               2011550.0
2               71         125000                 89800.0
3              710         540000                530980.0
4             7100          89000                 79280.0
5               84         205000                572770.0
14988  10032546000        1329400                     NaN
14989        10032       30570000               5502600.0
14990         1003       62000000               3720000.0
14991          100      200000000               3840000.0
14992           10         710000               2108700.0
14993      1003279        6750000                715330.0

s = df['Produtos'].astype(str)
i = pd.DataFrame(np.c_[s.str[0].astype(int), s.str.len()]).sort_values([0,1]).index
print (i)
Int64Index([10, 9, 8, 7, 11, 6, 0, 1, 2, 3, 4, 5], dtype='int64')
df = df.iloc[i]
print (df)
          Produtos  Estoque total  Valor Total de estoque
14992           10         710000               2108700.0
14991          100      200000000               3840000.0
14990         1003       62000000               3720000.0
14989        10032       30570000               5502600.0
14993      1003279        6750000                715330.0
14988  10032546000        1329400                     NaN
0               70          10000                  7180.0
1               70        2800000               2011550.0
2               71         125000                 89800.0
3              710         540000                530980.0
4             7100          89000                 79280.0
5               84         205000                572770.0