[Python Basics for AI] Pandas II
groupby



Hiearchical index : index 2개인 경우

reset_index() : index 새로 만들어 초기화
sort_values
-> hierarchical index일 때!




df.describe(), df.describe().T





Case study








pivot table, crosstab
pivot table




crosstab




Merge & Concat
Merge



join







concat


append : concat과 같은 효과

axis 주의
reset_index()
persistence


- Install
conda install openpyxl
conda install XlsxWriter

review
groupby
Hierarchical index
unstack()
swaplevel(), swaplevel.sortlevel(), swaplevel.sortvalues()
reset_index()
grouped
get_group()
aggregation : 요약된 통계정보
transformation : 개별 데이터 변환
filter : 특정 조건(boolean) 검색
.pivot_table()
pd.crosstab()
pd.merge(on=~), pd.merge(a,b,left_on=~, right_on)
- join
- INNER JOIN
- FULL JOIN
- LEFT JOIN
- RIGHT JOIN
-> pd.merge(df_a, df_b, on=~, how=~)
- index based :
pd.merge(df_a, df_b, right_index=True, left_index=True)
pd.concat(df_a, df_b, axis=~)
- persistence
- db connection
- XLS persistence
- Pickle persistence
.to_pickle()
.read_pickle()