If you want to get the information of the top 10 students in the total score, you may sort them according to the total score and then head(10)
do some operations, but what if the number of students in the top 10 is more than 10 swollen when you encounter the same ranking?
Today, let's take a look at the convenient function methods provided by pandas to make our data processing one step faster~
content:
1. Find the largest or smallest top N groups of data
When we process data, we often encounter a scenario, that is, to find the largest or smallest top N groups of data in this group of data . Under normal circumstances, we may use df.sort_values(columns, ascending=False).head(n)
to obtain, but there are often some data ranked in parallel that will be ruthlessly truncated and cannot be obtained. So, today we can try the following methods, and it will be fine.
Let's take the example of finding the largest first N groups of data as an example:
DataFrame.nlargest
( n,columns,keep='first')
Series.nlargest
( n=5,keep='first')
The optional value of the keep parameter: the default is first, optional last and all (literally)
We first construct a case data
>>> import pandas as pd
>>> df = pd.DataFrame({'population': [59000000, 65000000, 434000,
... 434000, 434000, 337000, 11300,
... 11300, 11300],
... 'GDP': [1937894, 2583560 , 12011, 4520, 12128,
... 17036, 182, 38, 311],
... 'alpha-2': ["IT", "FR", "MT", "MV", "BN",
... "IS", "NR", "TV", "AI"]},
... index=["Italy", "France", "Malta",
... "Maldives", "Brunei", "Iceland",
... "Nauru", "Tuvalu", "Anguilla"])
>>> df
population GDP alpha-2
Italy 59000000 1937894 IT
France 65000000 2583560 FR
Malta 434000 12011 MT
Maldives 434000 4520 MV
Brunei 434000 12128 BN
Iceland 337000 17036 IS
Nauru 11300 182 NR
Tuvalu 11300 38 TV
Anguilla 11300 311 AI
复制代码
For the above case data, if we want to get the top 3 sets of data with the most population fields, we find that the third place is 434000. If it is adopted head(3)
, there are actually 2 data rows that meet the requirements are missed by us; at this time df.nlargest(3, 'population',keep='all')
, we can get the results we need.
>>> df.head(3)
population GDP alpha-2
Italy 59000000 1937894 IT
France 65000000 2583560 FR
Malta 434000 12011 MT
>>> df.nlargest(3, 'population')
population GDP alpha-2
France 65000000 2583560 FR
Italy 59000000 1937894 IT
Malta 434000 12011 MT
# keep = 'all' 即表示满足排名的全部返回
>>> df.nlargest(3, 'population',keep='all')
population GDP alpha-2
France 65000000 2583560 FR
Italy 59000000 1937894 IT
Malta 434000 12011 MT
Maldives 434000 4520 MV
Brunei 434000 12128 BN
复制代码
Of course, we may face more complex requirements, such as taking the largest top N groups of data according to multiple fields. In this case, we need to take the top three data in GDP with the most population.
>>> df.nlargest(3, ['population', 'GDP'])
population GDP alpha-2
France 65000000 2583560 FR
Italy 59000000 1937894 IT
Brunei 434000 12128 BN
复制代码
For the smallest first N groups of data, the function is as follows (parameters have the same meaning):
DataFrame.nsmallest
( n, columns,keep='first')
Series.nsmallest
( n=5, keep='first')
2. 求当前元素和前一元素之间百分比变化
有时候,我们的数据可能是时间序列下的,为了更方便看到随着时间变化某行或列数据的变化率,这里就可以采用pct_change
方法直接获取。
pct_change
(periods=1,fill_method='pad', limit=None, freq=None, kwargs)
先看看对于Series
类型数据:
>>> s = pd.Series([90, 91, 85])
>>> s.pct_change()
0 NaN
1 0.011111
2 -0.065934
dtype: float64
# 设置间隔periods=2,在这里就是85相比90的变化率
>>> s.pct_change(periods=2)
0 NaN
1 NaN
2 -0.055556
dtype: float64
复制代码
对于有缺失值的情况,我们可以填充缺失值后参与计算或者在计算百分比时设置填充参数fill_method
:
>>> s = pd.Series([90, 91, None, 85])
>>> s
0 90.0
1 91.0
2 NaN
3 85.0
dtype: float64
>>> s.pct_change(fill_method='bfill')
0 NaN
1 0.011111
2 -0.065934
3 0.000000
dtype: float64
>>> s.pct_change(fill_method='ffill')
0 NaN
1 0.011111
2 0.000000
3 -0.065934
dtype: float64
>>> s.pct_change()
0 NaN
1 0.011111
2 0.000000
3 -0.065934
dtype: float64
复制代码
也可以直接对Dataframe
类型数据进行处理:
>>> df = pd.DataFrame({
... 'FR': [4.0405, 4.0963, 4.3149],
... 'GR': [1.7246, 1.7482, 1.8519],
... 'IT': [804.74, 810.01, 860.13]},
... index=['1980-01-01', '1980-02-01', '1980-03-01'])
>>> df
FR GR IT
1980-01-01 4.0405 1.7246 804.74
1980-02-01 4.0963 1.7482 810.01
1980-03-01 4.3149 1.8519 860.13
>>> df.pct_change()
FR GR IT
1980-01-01 NaN NaN NaN
1980-02-01 0.013810 0.013684 0.006549
1980-03-01 0.053365 0.059318 0.061876
>>> df = pd.DataFrame({
... '2016': [1769950, 30586265],
... '2015': [1500923, 40912316],
... '2014': [1371819, 41403351]},
... index=['GOOG', 'APPL'])
>>> df
2016 2015 2014
GOOG 1769950 1500923 1371819
APPL 30586265 40912316 41403351
# 对行进行操作axis=1或'columns'
>>> df.pct_change(axis = 1)
2016 2015 2014
GOOG NaN -0.151997 -0.086016
APPL NaN 0.337604 0.012002
>>> df.pct_change(axis = 'columns')
2016 2015 2014
GOOG NaN -0.151997 -0.086016
APPL NaN 0.337604 0.012002
复制代码
3. 将列表中每个元素转化为一行
有时候,我们的原始数据中某些元素可能是列表的形式,而我们需要对它进行展开操作,于是explode
方法就来了。
Series.explode
( ignore_index=False)
DataFrame.explode
( column, ignore_index=False)
先看看对Series
类型数据的处理:
>>> s = pd.Series([[1, 2, 3], 'foo', [], [3, 4]])
>>> s
0 [1, 2, 3]
1 foo
2 []
3 [3, 4]
dtype: object
# 默认情况下,会复制索引
>>> s.explode()
0 1
0 2
0 3
1 foo
2 NaN
3 3
3 4
dtype: object
# 设置参数ignore_index=True,则会重置索引
>>> s.explode(ignore_index=True)
0 1
1 2
2 3
3 foo
4 NaN
5 3
6 4
dtype: object
复制代码
再看看在Dataframe
类型数据下的操作:
>>> df = pd.DataFrame({'A': [[1, 2, 3], 'foo', [], [3, 4]], 'B': 1})
>>> df
A B
0 [1, 2, 3] 1
1 foo 1
2 [] 1
3 [3, 4] 1
# 默认情况下,会复制索引
>>> df.explode('A')
A B
0 1 1
0 2 1
0 3 1
1 foo 1
2 NaN 1
3 3 1
3 4 1
# 设置参数ignore_index=True,则会重置索引
>>> df.explode('A',ignore_index=True)
A B
0 1 1
1 2 1
2 3 1
3 foo 1
4 NaN 1
5 3 1
6 4 1
复制代码