pandas implements mysql windowing function group sorting operation (including null values)

pandas implements sorting operation within mysql window function group

row_number() over (partition by xx,yy order by mm asc ,nn desc ) as sos

# Import Data
import pandas as pd
df = pd.read_excel(r'C:\Users\Administrator\Desktop\test.xlsx')
df.head(15) 

1. Cumulative sorting within the group after grouping. The quantities are from high to low after grouping. The sorting is cumulative.

df1 = df.sort_values(["type", "brand", "topRanking", "quantity"], ascending=False)
df1.assign(rank=df1.assign(num=1).groupby(["type", "brand"], dropna=False)['num'].cumsum()) 

2. After grouping, the sorting within the group is sorted according to subcategories, and there is no continuous accumulation.

df1 = df.sort_values(["type", "brand", "topRanking", "quantity"], ascending=False)
df1["rank2"] =df1.groupby(["type", "brand", "topRanking"], dropna=False).quantity.rank(method="min", ascending=False,).astype("int")
df1