# Each step here potentially creates a new copy
df2 = df[df['sales'] > 1000]
df3 = df2.dropna()
df4 = df3.reset_index(drop=True)
df5 = df4[['sales', 'discount']]