def split_brand(brand_name_data):
return brand_name_data.split(' ')[-1]


user_df['brand_name'] = user_df.brand_info.apply(split_brand)




defining the function then applying it on the new column with existing column

user_df['brand_name'] = user_df.brand_info.apply(lambda brand_name: brand_name.split(' ')[-1]) with lambda applying

users_purchases = users_purchases.query('purchases >= 5') for querying

.agg({'brand_name': pd.Series.nunique}) finding out the unique value

users_purchases \
.merge(users_unique_brands, on='user_id') \
.merge(lovely_brand_purchases_df, on='user_id')

for merge purposes with primary key

ax = sns.barplot(x="lovely_brand", y="user_id", data=brands_loyalty) barplot

dropna() for dropping null values

df[
~(df['workingday'] == 0)
&(df['temp'] >= 10)
&(df['temp'] < 30)
] with working filtering

pd.set_option('display.max_rows', 85, index_col ="Column") for showing out the all indexes in table

df['ConvertedCp'].nlargest() finding out the max()

df.sort_values(by=['Country', 'ConvertedCp,'], ascending=[True, False], inplace = True) sorting advanced

df.drop(index = 4)

deleting with index

country_grp['Salary'].median().loc['Germany'] Germany country looking for the Salary

country_uses_python = country_grp['LanguageWorkedWith'].apply(lambda x: x.str.contains('Python').sum())

where Python word contains

df.columns = df.columns.str.replace('', '_')

df.columns = [x.upper() for x in df.columns]

df.apply(len, axis='columns')


df['email'].apply(len)

len(df['email'])

df['full_name'] = df['first'] + '' + df['last']

adding columsn

df.drop(columns = ['first', ' last'], inplace = True)

removing column

df[['first', 'last']] = df['full_name'].str.split(' ', expand = True)

mergin two columns as one

df.dropna(axis ='index', how ='all', subset=['last', 'email'])


droping null values with specfiying indexes