肝了3天，整理了50個Pandas高頻使用技巧，強(qiáng)烈建議收藏！

• DataFrame初印象
• 讀取表格型數(shù)據(jù)
• 篩選出特定的行
• 用pandas來繪圖
• 在DataFrame中新增行與列
• DataFrame中的統(tǒng)計分析與計算
• DataFrame中排序問題
• 合并多個表格
• 時序問題的處理
• 字符串類型數(shù)據(jù)的處理

DataFrame初印象

我們先來通過Python當(dāng)中的字典類型來創(chuàng)建一個DataFrame，

import?pandas?as?pd

data?=?{"Country":?["Canada",?"USA",?"UK"],
????????"Population":?[10.52*10**6,?350.1*10**6,?65.2*10**6]
???????}
df?=?pd.DataFrame(data)

df

??Country???Population
0??Canada???10520000.0
1?????USA??350100000.0
2??????UK???65200000.0

要是我們要獲取當(dāng)中的某一列，我們可以這么來做

df["Country"]

output

0????Portugal
1?????????USA
2??????France
Name:?Country,?dtype:?object

而當(dāng)我們想要獲取表格當(dāng)中每一列的數(shù)據(jù)格式的時候，可以這么做

df.dtypes

output

Country????????object
Population????float64
dtype:?object

讀取數(shù)據(jù)

import?pandas?as?pd
df?=?pd.read_csv("titanic.csv")

我們要是想要查看表格的前面幾行，可以這么做

df.head(7)

output

???PassengerId??Survived??Pclass??...?????Fare?Cabin??Embarked
0????????????1?????????0???????3??...???7.2500???NaN?????????S
1????????????2?????????1???????1??...??71.2833???C85?????????C
2????????????3?????????1???????3??...???7.9250???NaN?????????S
3????????????4?????????1???????1??...??53.1000??C123?????????S
4????????????5?????????0???????3??...???8.0500???NaN?????????S
5????????????6?????????0???????3??...???8.4583???NaN?????????Q
6????????????7?????????0???????1??...??51.8625???E46?????????S

df.tail(7)

output

?????PassengerId??Survived??Pclass??...????Fare?Cabin??Embarked
884??????????885?????????0???????3??...???7.050???NaN?????????S
885??????????886?????????0???????3??...??29.125???NaN?????????Q
886??????????887?????????0???????2??...??13.000???NaN?????????S
887??????????888?????????1???????1??...??30.000???B42?????????S
888??????????889?????????0???????3??...??23.450???NaN?????????S
889??????????890?????????1???????1??...??30.000??C148?????????C
890??????????891?????????0???????3??...???7.750???NaN?????????Q

要是遇到文件的格式是excel格式，pandas當(dāng)中也有相對應(yīng)的方法

df?=?pd.read_excel("titanic.xlsx")

可以通過pandas當(dāng)中的info()方法來獲取對表格數(shù)據(jù)的一個初步的印象

df.info()

output

'pandas.core.frame.DataFrame'>
RangeIndex:?891?entries,?0?to?890
Data?columns?(total?12?columns):
?#???Column???????Non-Null?Count??Dtype??
---??------???????--------------??-----??
?0???PassengerId??891?non-null????int64??
?1???Survived?????891?non-null????int64??
?2???Pclass???????891?non-null????int64??
?3???Name?????????891?non-null????object?
?4???Sex??????????891?non-null????object?
?5???Age??????????714?non-null????float64
?6???SibSp????????891?non-null????int64??
?7???Parch????????891?non-null????int64??
?8???Ticket???????891?non-null????object?
?9???Fare?????????891?non-null????float64
?10??Cabin????????204?non-null????object?
?11??Embarked?????889?non-null????object?
dtypes:?float64(2),?int64(5),?object(5)
memory?usage:?83.7+?KB

篩選出特定條件的行

要是我們想要篩選出年齡在30歲以上的乘客，我們可以這么來操作

df[df["Age"]?>?30]

output

?????PassengerId??Survived??Pclass??...?????Fare?Cabin??Embarked
1??????????????2?????????1???????1??...??71.2833???C85?????????C
3??????????????4?????????1???????1??...??53.1000??C123?????????S
4??????????????5?????????0???????3??...???8.0500???NaN?????????S
6??????????????7?????????0???????1??...??51.8625???E46?????????S
11????????????12?????????1???????1??...??26.5500??C103?????????S
..???????????...???????...?????...??...??????...???...???????...
873??????????874?????????0???????3??...???9.0000???NaN?????????S
879??????????880?????????1???????1??...??83.1583???C50?????????C
881??????????882?????????0???????3??...???7.8958???NaN?????????S
885??????????886?????????0???????3??...??29.1250???NaN?????????Q
890??????????891?????????0???????3??...???7.7500???NaN?????????Q

[305?rows?x?12?columns]

當(dāng)然我們也可以將若干個條件合起來，一同做篩選，例如

survived_under_45?=?df[(df["Survived"]==1)?&?(df["Age"]<45)]
survived_under_45

output

?????PassengerId??Survived??Pclass??...?????Fare?Cabin??Embarked
1??????????????2?????????1???????1??...??71.2833???C85?????????C
2??????????????3?????????1???????3??...???7.9250???NaN?????????S
3??????????????4?????????1???????1??...??53.1000??C123?????????S
8??????????????9?????????1???????3??...??11.1333???NaN?????????S
9?????????????10?????????1???????2??...??30.0708???NaN?????????C
..???????????...???????...?????...??...??????...???...???????...
874??????????875?????????1???????2??...??24.0000???NaN?????????C
875??????????876?????????1???????3??...???7.2250???NaN?????????C
880??????????881?????????1???????2??...??26.0000???NaN?????????S
887??????????888?????????1???????1??...??30.0000???B42?????????S
889??????????890?????????1???????1??...??30.0000??C148?????????C

[247?rows?x?12?columns]

df[df["Age"].notna()]

output

?????PassengerId??Survived??Pclass??...?????Fare?Cabin??Embarked
0??????????????1?????????0???????3??...???7.2500???NaN?????????S
1??????????????2?????????1???????1??...??71.2833???C85?????????C
2??????????????3?????????1???????3??...???7.9250???NaN?????????S
3??????????????4?????????1???????1??...??53.1000??C123?????????S
4??????????????5?????????0???????3??...???8.0500???NaN?????????S
..???????????...???????...?????...??...??????...???...???????...
885??????????886?????????0???????3??...??29.1250???NaN?????????Q
886??????????887?????????0???????2??...??13.0000???NaN?????????S
887??????????888?????????1???????1??...??30.0000???B42?????????S
889??????????890?????????1???????1??...??30.0000??C148?????????C
890??????????891?????????0???????3??...???7.7500???NaN?????????Q

[714?rows?x?12?columns]

df[df["Pclass"].isin([1,?2])]

output

?????PassengerId??Survived??Pclass??...?????Fare?Cabin??Embarked
1??????????????2?????????1???????1??...??71.2833???C85?????????C
3??????????????4?????????1???????1??...??53.1000??C123?????????S
6??????????????7?????????0???????1??...??51.8625???E46?????????S
9?????????????10?????????1???????2??...??30.0708???NaN?????????C
11????????????12?????????1???????1??...??26.5500??C103?????????S
..???????????...???????...?????...??...??????...???...???????...
880??????????881?????????1???????2??...??26.0000???NaN?????????S
883??????????884?????????0???????2??...??10.5000???NaN?????????S
886??????????887?????????0???????2??...??13.0000???NaN?????????S
887??????????888?????????1???????1??...??30.0000???B42?????????S
889??????????890?????????1???????1??...??30.0000??C148?????????C

[400?rows?x?12?columns]

df[(df["Pclass"]?==?1)?|?(df["Pclass"]?==?2)]

篩選出特定條件的行與列

df.loc[df["Age"]?>?40,"Name"]

output

6????????????????????????????????McCarthy,?Mr.?Timothy?J
11??????????????????????????????Bonnell,?Miss.?Elizabeth
15??????????????????????Hewlett,?Mrs.?(Mary?D?Kingcome)?
33?????????????????????????????????Wheadon,?Mr.?Edward?H
35????????????????????????Holverson,?Mr.?Alexander?Oskar
?????????????????????????????...????????????????????????
862????Swift,?Mrs.?Frederick?Joel?(Margaret?Welles?Ba...
865?????????????????????????????Bystrom,?Mrs.?(Karolina)
871?????Beckwith,?Mrs.?Richard?Leonard?(Sallie?Monypeny)
873??????????????????????????Vander?Cruyssen,?Mr.?Victor
879????????Potter,?Mrs.?Thomas?Jr?(Lily?Alexenia?Wilson)
Name:?Name,?Length:?150,?dtype:?object

df.loc[df["Age"]?>?40,["Name",?"Sex"]]

如果我們將逗號后面的部分直接用:來代替，則意味著要所有的列

df.loc[df["Age"]?>?40,:]

output

?????PassengerId??Survived??Pclass??...?????Fare?Cabin??Embarked
6??????????????7?????????0???????1??...??51.8625???E46?????????S
11????????????12?????????1???????1??...??26.5500??C103?????????S
15????????????16?????????1???????2??...??16.0000???NaN?????????S
33????????????34?????????0???????2??...??10.5000???NaN?????????S
35????????????36?????????0???????1??...??52.0000???NaN?????????S
..???????????...???????...?????...??...??????...???...???????...
862??????????863?????????1???????1??...??25.9292???D17?????????S
865??????????866?????????1???????2??...??13.0000???NaN?????????S
871??????????872?????????1???????1??...??52.5542???D35?????????S
873??????????874?????????0???????3??...???9.0000???NaN?????????S
879??????????880?????????1???????1??...??83.1583???C50?????????C

[150?rows?x?12?columns]

df.iloc[0:10,?0:3]

output

???PassengerId??Survived??Pclass
0????????????1?????????0???????3
1????????????2?????????1???????1
2????????????3?????????1???????3
3????????????4?????????1???????1
4????????????5?????????0???????3
5????????????6?????????0???????3
6????????????7?????????0???????1
7????????????8?????????0???????3
8????????????9?????????1???????3
9???????????10?????????1???????2

用Pandas來畫圖

我們還可以用Pandas來畫圖，而且實際用到的代碼量還比較的少

df.plot()

output

要是你想要單獨某一列的趨勢圖，我們也可以這么做

df["Age"].plot()

output

要是我們想要不同年齡對于船票費“Fare”的影響，畫圖可以這么來畫

df.plot.scatter(x?=?"Age",?y?=?"Fare",?alpha?=?0.6)

output

for?method_name?in?dir(df.plot):
????if?not?method_name.startswith("_"):
????????print(method_name)

output

area
bar
barh
box
density
hexbin
hist
kde
line
pie
scatter

df.plot.box()

output

要是我們希望可以分開來繪制圖形，就可以這么來操作

df.plot.area(figsize=(12,?4),?subplots=True)

output

要是我們想要將繪制好的圖片保存下來，可以直接使用savefig方法，

import?matplotlib.pyplot?as?plt
fig,?axs?=?plt.subplots(figsize=(12,?4))
df.plot.area(ax=axs)
fig.savefig("test.png")

output

如何新增一列

在DataFrame當(dāng)中新增一列其實不難，我們可以這么來操作

df["Date"]?=?pd.date_range("1912-04-02",?periods=len(df))
df.head()

output

???PassengerId??Survived??Pclass??...?Cabin?Embarked???????Date
0????????????1?????????0???????3??...???NaN????????S?1912-04-02
1????????????2?????????1???????1??...???C85????????C?1912-04-03
2????????????3?????????1???????3??...???NaN????????S?1912-04-04
3????????????4?????????1???????1??...??C123????????S?1912-04-05
4????????????5?????????0???????3??...???NaN????????S?1912-04-06

[5?rows?x?13?columns]

def?define_age(age):
????if?age?????????return?"少年"
????elif?age?>=?18?and?age?????????return?"青年"
????elif?age?>=?35?and?age?????????return?"中年"
????else:
????????return?"老年"

然后再用apply來實現(xiàn)

df["Generation"]?=?df["Age"].apply(define_age)
df.head()

output

???PassengerId??Survived??Pclass??...?Cabin?Embarked??Generation
0????????????1?????????0???????3??...???NaN????????S??????????青年
1????????????2?????????1???????1??...???C85????????C??????????中年
2????????????3?????????1???????3??...???NaN????????S??????????青年
3????????????4?????????1???????1??...??C123????????S??????????中年
4????????????5?????????0???????3??...???NaN????????S??????????中年

[5?rows?x?13?columns]

如果我們想給表格中的列名重新命名的話，可以使用rename方法，

df_renamed?=?df.rename(columns={"Name":"Full?Name",?"Sex":?"Gender",?"Ticket":?"FareTicket"})
df_renamed.head()

output

DataFrame中的統(tǒng)計分析

在Pandas中也提供了很多相關(guān)的方法來進(jìn)行數(shù)據(jù)的統(tǒng)計分析

print(df["Age"].mean())
print(df["Age"].max())
print(df["Age"].min())
print(df["Age"].median())

29.69911764705882
80.0
0.42
28.0

同時我們也可以使用describe()方法

df.describe()

output

???????PassengerId????Survived??????Pclass??...???????SibSp???????Parch????????Fare
count???891.000000??891.000000??891.000000??...??891.000000??891.000000??891.000000
mean????446.000000????0.383838????2.308642??...????0.523008????0.381594???32.204208
std?????257.353842????0.486592????0.836071??...????1.102743????0.806057???49.693429
min???????1.000000????0.000000????1.000000??...????0.000000????0.000000????0.000000
25%?????223.500000????0.000000????2.000000??...????0.000000????0.000000????7.910400
50%?????446.000000????0.000000????3.000000??...????0.000000????0.000000???14.454200
75%?????668.500000????1.000000????3.000000??...????1.000000????0.000000???31.000000
max?????891.000000????1.000000????3.000000??...????8.000000????6.000000??512.329200

[8?rows?x?7?columns]

當(dāng)然我們也可以對于特定幾列的數(shù)據(jù)進(jìn)行統(tǒng)計分析

df.agg(
????{
????????"Age":?["min",?"max",?"mean"],
????????"Fare":?["min",?"max",?"mean"]
????}
)

output

????????????Age????????Fare
min????0.420000????0.000000
max???80.000000??512.329200
mean??29.699118???32.204208

df[["Sex",?"Age"]].groupby("Sex").mean()

output

??????????????Age
Sex??????????????
female??27.915709
male????30.726645

另外，value_counts()方法也可以針對單獨某一列數(shù)據(jù)，看一下數(shù)據(jù)的具體分布，

df["Pclass"].value_counts()

output

3????491
1????216
2????184
Name:?Pclass,?dtype:?int64

DataFrame中的排序問題

我們假設(shè)有這么一組數(shù)據(jù)，

data?=?{
????"Name":?["Mike",?"Peter",?"Clara",?"Tony",?"John"],
????"Age":?[30,?26,?20,?22,?25]
}
df?=?pd.DataFrame(data)
df

output

????Name??Age
0???Mike???30
1??Peter???26
2??Clara???20
3???Tony???22
4???John???25

我們可以將數(shù)據(jù)按照“Age”年齡這一列來進(jìn)行排序

df.sort_values(by="Age")

output

????Name??Age
2??Clara???20
3???Tony???22
4???John???25
1??Peter???26
0???Mike???30

當(dāng)然我們也可以按照降序來進(jìn)行排列

df.sort_values("Age",?ascending=False)

output

????Name??Age
0???Mike???30
1??Peter???26
4???John???25
3???Tony???22
2??Clara???20

合并多個表格

例如我們有這么兩個表格，

df1?=?pd.DataFrame(
????{
????????"Name":?["Mike",?"John",?"Clara",?"Linda"],
?????????"Age":?[30,?26,?20,?22]
????}
)

df2?=?pd.DataFrame(
?????{
?????????"Name":?["Brian",?"Mary"],
?????????"Age":?[45,?38]
?????}
)
df_names_ages?=?pd.concat([df1,?df2],?axis=0)
df_names_ages

output

????Name??Age
0???Mike???30
1???John???26
2??Clara???20
3??Linda???22
0??Brian???45
1???Mary???38

df1?=?pd.DataFrame(
????{
????????"Name":?["Mike",?"John",?"Clara",?"Sara"],
?????????"Age":?[30,?26,?20,?22],
????????"City":?["New?York",?"Shanghai",?"London",?"Paris"],
????}
)

df2?=?pd.DataFrame(
?????{
?????????"City":?["New?York",?"Shanghai",?"London",?"Paris"],
?????????"Occupation":?["Machine?Learning?Enginner",?"Data?Scientist",?"Doctor","Teacher"]
?????}
)

df_merged?=?pd.merge(df1,df2,how="left",?on="City")
df_merged

output

????Name??Age??????City?????????????????Occupation
0???Mike???30??New?York??Machine?Learning?Enginner
1???John???26??Shanghai?????????????Data?Scientist
2??Clara???20????London?????????????????????Doctor
3???Sara???22?????Paris????????????????????Teacher

時序問題的處理

在時序問題的處理上，小編之前專門寫過一篇文章，具體可以看

干貨分享 | Pandas處理時間序列的數(shù)據(jù)

例如我們有這么一個數(shù)據(jù)集

df?=?pd.read_csv("air_quality.csv")
df?=?df.rename(columns={"date.utc":?"datetime"})
df.head()

output

????city?country???????????????????datetime?location?parameter??value
0??Paris??????FR??2019-06-21?00:00:00+00:00??FR04014???????no2???20.0
1??Paris??????FR??2019-06-20?23:00:00+00:00??FR04014???????no2???21.8
2??Paris??????FR??2019-06-20?22:00:00+00:00??FR04014???????no2???26.5
3??Paris??????FR??2019-06-20?21:00:00+00:00??FR04014???????no2???24.9
4??Paris??????FR??2019-06-20?20:00:00+00:00??FR04014???????no2???21.4

我們看一下目前“datetime”這一列的數(shù)據(jù)類型

df.dtypes

output

city??????????object
country???????object
datetime??????object
location??????object
parameter?????object
value????????float64
dtype:?object

df["datetime"]?=?pd.to_datetime(df["datetime"])
df["datetime"].head()

output

0???2019-06-21?00:00:00+00:00
1???2019-06-20?23:00:00+00:00
2???2019-06-20?22:00:00+00:00
3???2019-06-20?21:00:00+00:00
4???2019-06-20?20:00:00+00:00
Name:?datetime,?dtype:?datetime64[ns,?UTC]

我們便可以查看起始的日期

df["datetime"].min(),?df["datetime"].max()

output

(Timestamp('2019-05-07?01:00:00+0000',?tz='UTC'),
?Timestamp('2019-06-21?00:00:00+0000',?tz='UTC'))

中間相隔的時間

df["datetime"].max()?-?df["datetime"].min()

output

Timedelta('44?days?23:00:00')

文本數(shù)據(jù)的處理問題

當(dāng)我們的數(shù)據(jù)集中存在文本數(shù)據(jù)時，pandas內(nèi)部也有相對應(yīng)的處理方法

data?=?{"Full?Name":?["Peter?Parker",?"Linda?Elisabeth",?"Bob?Dylan"],
????????"Age":?[40,?50,?60]}

df?=?pd.DataFrame(data)
df

output

?????????Full?Name??Age
0?????Peter?Parker???40
1??Linda?Elisabeth???50
2????????Bob?Dylan???60

可以用str方法將這些文本數(shù)據(jù)摘取出來，然后再進(jìn)一步操作

df["Full?Name"].str.lower()

output

0?????? peter parker
1????linda?elisabeth
2??????????bob?dylan
Name:?Full?Name,?dtype:?object

或者也可以這樣來操作

df["Last?Name"]?=?df["Full?Name"].str.split("?").str.get(-1)
df

output

?????????Full?Name??Age??Last?Name
0?????Peter?Parker???40?????Parker
1??Linda?Elisabeth???50??Elisabeth
2????????Bob?Dylan???60??????Dylan

這樣我們可以將其“姓”的部分給提取出來，同樣的我們也可以提取“名”的部分

df["First?Name"]?=?df["Full?Name"].str.split("?").str.get(0)
df

output

?????????Full?Name??Age??Last?Name?First?Name
0?????Peter?Parker???40?????Parker??????Peter
1??Linda?Elisabeth???50??Elisabeth??????Linda
2????????Bob?Dylan???60??????Dylan????????Bob

我們也可以通過contains方法來查看字段中是不是包含了某一個字符串

df["Full?Name"].str.contains("Bob")

output

0????False
1????False
2?????True
Name:?Full?Name,?dtype:?bool

同樣也是通過str方法將文本數(shù)據(jù)也提取出來再進(jìn)行進(jìn)一步的操作