Here we have the table with the number of wine bottles by countries and varieties:
df
Output:
+----+--------------------------+-------------+-------------+----------+----------+---------+---------+------+
| | variety | Argentina | Australia | Canada | France | Italy | Spain | US |
|----+--------------------------+-------------+-------------+----------+----------+---------+---------+------|
| 0 | Bordeaux-style Red Blend | 79 | 28 | 15 | 862 | 2 | 6 | 1235 |
| 1 | Cabernet Sauvignon | 391 | 155 | 3 | 30 | 29 | 42 | 2463 |
| 2 | Chardonnay | 204 | 206 | 14 | 1587 | 110 | 37 | 2778 |
| 3 | Merlot | 34 | 7 | 4 | 67 | 47 | 11 | 703 |
| 4 | Pinot Noir | 68 | 72 | 14 | 1121 | 1 | 11 | 5449 |
| 5 | Red Blend | 209 | 60 | 5 | 218 | 1605 | 683 | 1985 |
| 6 | Rosé | 30 | 18 | 2 | 1120 | 4 | 106 | 510 |
| 7 | Sauvignon Blanc | 52 | 19 | 4 | 616 | 18 | 16 | 669 |
| 8 | Syrah | 46 | 15 | 9 | 110 | 49 | 27 | 1728 |
| 9 | White Blend | 30 | 12 | 5 | 167 | 295 | 236 | 410 |
+----+--------------------------+-------------+-------------+----------+----------+---------+---------+------+
You have to reshape into a long format (3 columns) and save the result to df_long.
Tip: At the end you should get the dataset with columns containing varieties, countries and its values. Try melt() function.