Сохранение всех строк с верхними n элементами на основе количества значений в DataFrame

Вам нужно:

#convert column to numeric
df['Price'] = df['Price'].str.replace(',','.').astype(float)

#get top2 values from index
print (df['Price'].value_counts().iloc[:2])
78.0     3
199.0    3
Name: Price, dtype: int64

#filter rows with top2 values (78, 199)
df = df[df['Price'].isin(df['Price'].value_counts().iloc[:2].index)]
print (df)
  InvoiceId StockCode  Price
0       XXX  ProductA  199.0
1       XXX  ProductB   78.0
3       YYY  ProductB   78.0
4       YYY  ProductA  199.0
5       ZZZ  ProductA  199.0
6       ZZZ  ProductB   78.0

#count top2
df1 = pd.crosstab(df['InvoiceId'],
                  df['StockCode'])
print (df1)
StockCode  ProductA  ProductB
InvoiceId                    
XXX               1         1
YYY               1         1
ZZZ               1         1