I have a problem and need to solve it using Pandas/Python. Not sure how to achieve it and would be great if someone help here to build the logic.

I have to generate the output file as below:

df = pd.DataFrame({'priority': [1, 1, 1, 2, 2, 3],'db_name': ['corp', 'corp', 'corp', 'sales', 'sales', 'market'],'tbl_name': ['c_tbl1', 'c_tbl1', 'c_tbl1', 's_tbl1', 's_tbl2', 'm_tbl1'],'partition': ['202301', '202302', '202303', '202301', '202302', '202301'],'size_gb': [5, 5, 10, 1, 2, 3]})

Logic would be like this for priority 1 - three entries presents with different sizes, if the size is 10 GB single entry in output file with t_size = XL or sump of the size create single entry with t_size = XL similarly for other priorities size is less than 3 GB then t_size = S otherwise M.

I tried to loop using Pandas data frames, couldn't proceed as I am not proficient in Python.

Try it:

df = pd.read_csv("F:\\sales_data.csv", header=0)df = pd.DataFrame(df.pivot_table(index=['priority', 'db_name', 'tbl_name'],values=['partition', 'size_gb'],aggfunc={'partition': lambda x: ";".join(str(v) for v in x), 'size_gb': 'sum'}).to_records())
df['t_size'] = np.where(df['size_gb'] >= 10, 'XL',np.where(df['size_gb'] < 3, 'S', 'M'))df.drop(columns=['size_gb'], inplace=True)
print(df)

Output:

  priority  db_name tbl_name    partition   t_size
0   1        corp   c_tbl1    202301;202302  XL
1   1        corp   c_tbl1    202303         XL
2   2        sales  s_tbl1    202301         S
3   2        sales  s_tbl2    202302         S
4   3        market m_tbl1    202301         M