Group data sets and apply a function to each group （ Whether it's aggregation or transformation ）, It is usually an important part of data analysis . Loading the dataset 、 The fusion 、 When you're ready , It is usually used to calculate group statistics or generate pivot tables .pandas Provides a flexible and efficient gruopby function , It allows you to slice data sets in a natural way 、 cutting 、 Abstract and so on .

Relational databases and SQL（Structured Query Language, Structured query language ） One of the reasons why it is so popular is that it can easily connect data 、 Filter 、 Transformation and aggregation . however , image SQL The types of grouping operations that such query languages can perform are very limited . In this chapter you will see , because Python and pandas Strong expression skills , We can perform much more complex grouping operations （ Use whatever is acceptable pandas Object or NumPy Array functions ）. In this chapter , You will learn ：

• Use one or more keys （ The form can be a function 、 An array or DataFrame Name ） Division pandas object .
• Calculate overview statistics for groups , Like quantity 、 Mean or standard deviation , Or user-defined functions .
• Apply intra group conversions or other operations , Normalization 、 Linear regression 、 Ranking or selecting subsets, etc .
• Calculate pivot table or crosstab .
• Perform quantile analysis and other statistical grouping analysis .

note ： Aggregation of time series data （groupby One of the special uses of ） Also called resampling （resampling）, This book will be in 11 It is explained separately in chapter .

10.1 GroupBy Mechanism

Hadley Wickham（ Many popular R Author of the language pack ） Created a term for grouping operations "split-apply-combine"（ Split － application －