coding►c++

c++ compiler

指定编译器

1
2
3
4
5
export PATH=/opt/scylladb/bin:$PATH
export LD_LIBRARY_PATH=/opt/scylladb/lib64:$LD_LIBRARY_PATH
export LD_LIBRARY_PATH=/opt/scylladb/lib:$LD_LIBRARY_PATH
export CC=/opt/scylladb/bin/gcc
export CXX=/opt/scylladb/bin/g++

编译选项

1
2
3
4
5
6
# warning
-Wall			开启警告（all: 应该开启的最小警告集合）
-Wextra		开启扩展警告
-Werror 	所有警告视为错误
-Wno-error=...	关闭某项警告视为错误
-Wno-<...>			关闭某项警告视为错误

附录

-Wall

1
2
3
4
5
6
7
8
9
-Waddress -Warray-bounds (only with -O2) -Wc++0x-compat -Wchar-subscripts
-Wenum-compare (in C/Objc; this is on by default in C++) -Wimplicit-int (C and
 Objective-C only) -Wimplicit-function-declaration (C and Objective-C only) 
-Wcomment -Wformat -Wmain (only for C/ObjC and unless -ffreestanding) 
-Wmissing-braces -Wnonnull -Wparentheses -Wpointer-sign -Wreorder -Wreturn-type 
-Wsequence-point -Wsign-compare (only in C++) -Wstrict-aliasing 
-Wstrict-overflow=1 -Wswitch -Wtrigraphs -Wuninitialized -Wunknown-pragmas 
-Wunused-function -Wunused-label -Wunused-value -Wunused-variable 
-Wvolatile-register-var

-Wextra

1
2
3
4
-Wclobbered -Wempty-body -Wignored-qualifiers -Wmissing-field-initializers
-Wmissing-parameter-type (C only) -Wold-style-declaration (C only) -Woverride-init
-Wsign-compare -Wtype-limits -Wuninitialized -Wunused-parameter (only with -Wunused
 or -Wall) -Wunused-but-set-parameter (only with -Wunused or -Wall)

warn 示例配置

1
-Wall -Wextra -Waggregate-return -Wcast-align -Wcast-qual -Wdisabled-optimization -Wdiv-by-zero -Wendif-labels -Wformat-extra-args -Wformat-nonliteral -Wformat-security -Wformat-y2k -Wimplicit -Wimport -Winit-self -Winline -Winvalid-pch -Wjump-misses-init -Wlogical-op -Werror=missing-braces -Wmissing-declarations -Wno-missing-format-attribute -Wmissing-include-dirs -Wmultichar -Wpacked -Wpointer-arith -Wreturn-type -Wsequence-point -Wsign-compare -Wstrict-aliasing -Wstrict-aliasing=2 -Wswitch -Wswitch-default -Werror=undef -Wno-unused -Wvariadic-macros -Wwrite-strings -Wc++-compat -Werror=declaration-after-statement -Werror=implicit-function-declaration -Wmissing-prototypes -Werror=nested-externs -Werror=old-style-definition -Werror=strict-prototypes
coding►c++

内存泄漏排查

简述

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
$ pid=$(pgrep ranker_service)
$ pmap -x $pid
9085:   ./bin/ranker_service --cloud_region=aws-se --gray_cloud_list=none
Address           Kbytes     RSS   Dirty Mode  Mapping
0000000000400000   19180    7236       0 r-x-- ranker_service
00000000018bb000     280     168      24 r---- ranker_service
0000000001901000      92      40      40 rw--- ranker_service
0000000001918000     212     120     120 rw---   [ anon ]
0000023412800000  262144  216280  216280 rw---   [ anon ]
000002342f000000  528384  524292  524292 rw---   [ anon ]
0000023453800000  262144  262144  262144 rw---   [ anon ]
0000023469c00000  262144  262144  262144 rw---   [ anon ]
0000023486000000  262144  262144  262144 rw---   [ anon ]
00000234aec00000  524288  518296  518296 rw---   [ anon ]
00000234db400000  262144  150220  150220 rw---   [ anon ]
0000023546400000 23330816 23188160 23188160 rw---   [ anon ]
0000023b50400000  303104  294920  294920 rw---   [ anon ]
0000023b69000000  475136  458768  458768 rw---   [ anon ]
...
$ gcore $pid

dump

1
2
3
4
5
6
7
8
$ gdb bin/ranker_service core.9085
(gdb) dump binary memory result.bin 0x23546400000 0x23586400000  # dump 1G 内存数据

## dump 内存块 0x23546400000 全部数据
# 计算地址
> hex(0x0000023546400000 + 23330816 * 1024) # python3
'0x23ad6400000'
(gdb) dump binary memory result.bin 0x23546400000 0x23ad6400000  # dump 区块全部数据 23330816KB

统计 vtable

1
2
$ hexdump result.bin | awk '{printf "%s%s%s%s\n%s%s%s%s\n", $5,$4,$3,$2,$9,$8,$7,$6}' | sort | uniq -c | sort -nr  > hex.t
$ hexdump core.9085 | awk '{printf "%s%s%s%s\n%s%s%s%s\n", $5,$4,$3,$2,$9,$8,$7,$6}' | sort | uniq -c | sort -nr  | head

image-20220613180006614

image-20220613180039586

参考

linux►software

nginx creat self-signed SSL certificate

创建证书

1
2
3
4
5
6
7
8
9
10
11
12
13
14
# 自认证证书
# 生成 nginx-selfsigned.pass.key
$ openssl genrsa -des3 -passout pass:over4chars -out nginx-selfsigned.pass.key 2048
# 生成 nginx-selfsigned.key
$ openssl rsa -passin pass:over4chars -in nginx-selfsigned.pass.key -out nginx-selfsigned.key
$ rm nginx-selfsigned.pass.key # 可以删除了
# 生成 nginx-selfsigned.csr
$ openssl req -new -key nginx-selfsigned.key -out nginx-selfsigned.csr # 一直回车
# 生成 nginx-selfsigned.crt
$ openssl x509 -req -sha256 -days 365 -in nginx-selfsigned.csr -signkey nginx-selfsigned.key -out nginx-selfsigned.crt

# 需要的文件
- nginx-selfsigned.key
- nginx-selfsigned.crt

或者

1
$ openssl req -x509 -nodes -days 365 -newkey rsa:2048 -keyout /etc/ssl/private/nginx-selfsigned.key -out /etc/ssl/certs/nginx-selfsigned.crt

设置 nginx 使用 ssl

把生成的整数文件放到 /certs 目录下。

1
2
3
$ sudo mkdir /certs
$ sudo mv nginx-selfsigned.key /certs
$ sudo mv nginx-selfsigned.crt /certs

生成 dhparam.pem 文件。

1
$ openssl dhparam -out /certs/dhparam.pem 4096

创建 /etc/nginx/snippets/self-signed.conf ，输入如下内容。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
ssl_certificate /certs/nginx-selfsigned.crt;
ssl_certificate_key /certs/nginx-selfsigned.key;

ssl_protocols TLSv1 TLSv1.1 TLSv1.2;
ssl_prefer_server_ciphers on;
ssl_ciphers "EECDH+AESGCM:EDH+AESGCM:AES256+EECDH:AES256+EDH";
ssl_ecdh_curve secp384r1;
ssl_session_cache shared:SSL:10m;
ssl_session_tickets off;
ssl_stapling on;
ssl_stapling_verify on;
resolver 8.8.8.8 8.8.4.4 valid=300s;
resolver_timeout 5s;
add_header Strict-Transport-Security "max-age=63072000; includeSubdomains";
add_header X-Frame-Options DENY;
add_header X-Content-Type-Options nosniff;

ssl_dhparam /certs/dhparam.pem;

设置 nginx site 配置。

1
2
3
4
5
6
7
8
9
10
server {
    listen 80 default_server;
    listen [::]:80 default_server;
    listen 443 ssl http2 default_server;
    listen [::]:443 ssl http2 default_server;

    server_name server_domain_or_IP;
    include snippets/self-signed.conf;
    ...
}

配置 http 跳转 https

按需配置。

1
2
3
4
5
6
server {
    listen 80 default_server;
    listen [::]:80 default_server;
    server_name server_domain_or_IP;
    return 302 https://$server_name$request_uri;
}

参考

coding►c++►tools►gperftools

gperftools

安装

1
$ yum install gperftools gperftools-devel -y

Heap Profile

1
2
3
4
5
6
7
8
# dump
$ env LD_PRELOAD="/usr/lib64/libtcmalloc_and_profiler.so" HEAPPROFILE="heap-profile" <binary>
$ env LD_PRELOAD="/usr/lib64/libtcmalloc.so" HEAPPROFILE="heap-profile" <binary>
$ env LD_PRELOAD="/usr/lib64/libtcmalloc.so" HEAPPROFILE="/tmp/profile" HEAP_PROFILE_TIME_INTERVAL=60 <binary>

# analysis
$ pprof --text <binary> /tmp/profile....heap
$ pprof --svg <binary> /tmp/profile....heap > heap.svg
架构►container►k8s

k8s usage

[toc]

节点管理

删除节点

1
$ sudo kubeadm reset cleanup-node  # master 节点也可以清除

标签

1
2
3
4
5
# 查看 node 标签
$ kubectl get nodes --show-labels

# 打标签
$ kubectl label nodes <your-node-name> disktype=ssd

设置 pod 的 node 标签选择（nodeSelector）

1
2
3
4
5
6
7
8
9
10
11
12
13
apiVersion: v1
kind: Pod
metadata:
  name: nginx
  labels:
    env: test
spec:
  containers:
  - name: nginx
    image: nginx
    imagePullPolicy: IfNotPresent
  nodeSelector:
    disktype: ssd

设置 deployment node 标签选择

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
spec:
	template:
    # 按标签选择机器
    nodeSelector:
      <label-name>: <label-value>
    # 设置机器亲和性
    affinity:
      # 设置 node 亲和性
      nodeAffinity:
        requiredDuringSchedulingIgnoredDuringExecution:
          nodeSelectorTerms:
            - matchExpressions:
                - key: <label-name>
                  operator: In
                  values:
                    - <label-value>
      # 设置 pod 亲和性
      podAntiAffinity:
        requiredDuringSchedulingIgnoredDuringExecution:
          - labelSelector:
              matchExpressions:
                - key: app.kubernetes.io/name
                  operator: In
                  values:
                    - <pod-name>
                    - <pod-name>
            topologyKey: kubernetes.io/hostname

Deployment 管理

1
kubectl delete --all deployments d-n=<ns>

Pod 管理

1
2
3
4
5
6
7
8
9
10
# 查看 pod  信息
$ kubectl get pods -A  # A = all-namespaces

# 删除 pod
$ kubectl delete pod <name> --namespace=<ns>
# 批量删除
kubectl delete pod --all -n=<ns>

# 查看 pod 详情; ip 等
$ kubectl describe pod <name> --namespace=<ns>

Service

1
2
3
4
5
# 列出所有 service
k get services -o wide -A --sort-by=.metadata.name 

# 获取 service 详情; nodeport 等
k describe service kubernetes-dashboard -n kube-system

Token

1
microk8s kubectl create token -n kube-system default --duration=8544h

私有仓库

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
# docker login
docker login hub.company.com

# 不指定 namespace, 设置 default
kubectl create secret generic regcred \
    --from-file=.dockerconfigjson=~/.docker/config.json \
    --type=kubernetes.io/dockerconfigjson

# 每个 namespace 需要单独设置
kubectl create secret generic regcred \
    --from-file=.dockerconfigjson=~/.docker/config.json \
    --type=kubernetes.io/dockerconfigjson \
    --namespace=ace

# 修改 deployment
apiVersion: v1
kind: Pod
metadata:
  name: private-reg
spec:
  containers:
  - name: private-reg-container
    image: <your-private-image>
  # 添加 imagePullSecrets
  imagePullSecrets:
  - name: regcred

Secret

1
2
3
4
5
6
7
8
9
10
11
12
# 创建
kubectl create secret generic regcred \
    --from-file=.dockerconfigjson=$HOME/.docker/config.json \
    --type=kubernetes.io/dockerconfigjson \
    --namespace=default

# 查看
kubectl get secrets
kubectl get secretss <secret> --output=yaml

# 删除
kubectl delete secret <secret>

Namespace

1
2
3
4
5
# 列出所有的 namespace
$ kubectl get namespaces

# 创建命名空间
$ kubectl create namespace <space-name>

登录容器

1
kubectl exec --stdin --tty <pod-instance> -- /bin/bash

工具

  • k9s，基于 SHELL 的 K8S 客户端工具
  • k8s Lens，桌面端 K8S 客户端工具
python►library►matplotlib

matplotlib

参数

alpha

  • 透明度
  • float [0, 1]

bins

  • 分段区间
  • 可以是一个整数或一个序列
  • 如果是整数n，将会均匀分为n个区间，否则按指定的序列进行分段

配置

subplot

绘制子图。

原型

1
subplot(nrows, ncols, index, **kwargs)

示例

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
import matplotlib.pyplot as plt

y = [i for i in range(100)]
z = [i * 1.1 for i in range(100)]

plt.subplot(2, 2, 1)
plt.xlabel('x label')
plt.ylabel('y label')
plt.plot(y, label='y')
plt.plot(z, label='z')
plt.legend()

#### no legend
plt.subplot(2, 2, 2)
plt.xlabel('x label')
plt.ylabel('y label')
plt.plot(y, label='y')
plt.plot(z, label='z')

plt.subplot(2, 2, 4)
plt.xlabel('x label')
plt.ylabel('y label')
plt.plot(y, label='y')
plt.plot(z, label='z')
plt.legend()

# no NO.3 subplot 
plt.show()

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
# 数据来自鸢尾花数据集，绘制散点图
columns = ['Id', 'SepalLengthCm', 'SepalWidthCm', 'PetalLengthCm', 'PetalWidthCm', 'Label']
for i in range(2, 5):
    plt.subplot(1, 3, i - 1)
    plt.scatter(features[:, 1], features[:, i], c=labels, s=50, cmap='viridis')
    plt.title(columns[i])
plt.show()

# 自定义 subplot size
start = 2
end = 5
fg, axs = plt.subplots(end - start, end - start, figsize=(25, 25))
for i in range(start, end):
    for j in range(start, end):
        axs[i - start, j - start].scatter(features[:, 1], features[:, i], c=labels, s=25, cmap='viridis')
        axs[i - start, j - start].set_title(columns[i] + ' - ' + columns[j])
python►library►matplotlib

matplotlib usage

简介

Matplotlib

  • 数据可视化

安装

1
2
3
sudo pip3 install matplotlib
# or
sudo apt install python3-matplotlib

示例程序

sample 1

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
import numpy as np
import matplotlib.pyplot as plt


x = np.linspace(0, 10, 1000)    # 作图的变量自变量
y = np.sin(x) + 1               # 因变量y
z = np.cos(x**2) + 1            # 因变量z

plt.figure(figsize=(8, 4))      # 设置图像大小
plt.plot(x, y, label='$\sin x+1$', color='red', linewidth=2)    # 作图, 设置标签, 线条颜色, 线条大小
plt.plot(x, z, 'b--', label='$\cos x^2+1$')
plt.xlabel('Time(s) ')  # x轴名称
plt.ylabel('Volt')      # y轴名称
plt.ylim(0, 2.2)        # 显示的y轴范围
plt.legend()            # 显示图例
plt.show()              # 显示作图结果

运行结果

Sample 2 - subplot

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
#!/bin/python
import matplotlib.pyplot as plt

# add a subplot
ax1=plt.subplot(2, 2, 1)
# add a subplot with no frame
ax2=plt.subplot(2, 2, 2, frameon=False)
# add a polar subplot
plt.subplot(2, 2, 3, projection='polar')
# add a red subplot that shares the x-axis with ax1
plt.subplot(2, 2, 4, sharex=ax1, facecolor='red')

#delete ax2 from the figure
plt.delaxes(ax2)
#add ax2 to the figure again
plt.subplot(ax2)

plt.show()

Sample 3 - drow

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
from cycler import cycler
import numpy as np
import matplotlib.pyplot as plt


x = np.linspace(0, 2 * np.pi)
offsets = np.linspace(0, 2*np.pi, 4, endpoint=False)
# Create array with shifted-sine curve along each column
yy = np.transpose([np.sin(x + phi) for phi in offsets])

# 1. Setting prop cycle on default rc parameter
plt.rc('lines', linewidth=4)
plt.rc('axes', prop_cycle=(cycler('color', ['r', 'g', 'b', 'y']) +
                           cycler('linestyle', ['-', '--', ':', '-.'])))
fig, (ax0, ax1) = plt.subplots(nrows=2)
ax0.plot(yy)
ax0.set_title('Set default color cycle to rgby')

# 2. Define prop cycle for single set of axes	color & width
ax1.set_prop_cycle(cycler('color', ['c', 'm', 'y', 'k']) +
                   cycler('lw', [1, 2, 3, 4]))	
ax1.plot(yy)
ax1.set_title('Set axes color cycle to cmyk')

# Tweak spacing between subplots to prevent labels from overlapping
fig.subplots_adjust(hspace=0.3)
plt.show()

Sample 4 - set size && save picture

1
2
3
4
5
6
#### way 1
from matplotlib.pyplot import figure
figure(figsize=(20, 10), dpi=80,)

#### way 2
plt.figure(figsize=(15, 10), dpi=80)
python►library►pandas

pandas

[toc]

简介

  • 强大的数据分析和探索工具
  • 包含高级的数据结构和精巧的工具
  • 构建在Numpy之上
  • 需要安装xlrd和xlwt支持excel文件读写

数据结构

Series

  • 序列, 类似于一维数组

获取值

1
s.values

DataFrame

  • 相当于一张二维的表格, 类似二维数组
  • 每一列都是一个Series

处理CSV文件

读取文件

1
2
# 读取csv文件，并转化为DataFrame
df = pd.read_csv('../data/train.csv')

导出文件

1
df.to_csv("output.csv", index=False)

DataFrame方法

数据探索

读取列

1
print(data.columns)

选取数据

  • iloc
  • DF.clomun_name : 通过列名选择

打印列的类型

1
df['Column'].dtype

过滤行

1
df['Column'][df['Column'].isna()]

转换类型

1
df['Column'] = df['Column'].astype(str)

NaN 值处理

包含 NaN 的列

1
2
for c in df.loc[:, df.isna().any()]:
  pass

打印每列包含 NaN 的列数

1
df.isna().sum()

填充 NaN 值

1
df['Column'] = df['Column'].fillna(df['Column'].mean())  # 是用 Column 的平均值填充该列的 NaN 值

其他

iloc

从DataFrame中选择数据

1
2
3
df.iloc[:]		# 选择所有行，所有列
df.iloc[:, :4]	# 选择所有行，前四列
df.iloc[[1, 2], [0, 1]]	# 选择第2,3 行， 第1,2列

示例程序

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
### With Series
bovenson@ThinkCentre:~/Git/notes/Python/Pandas$ python3
Python 3.5.3 (default, Jan 19 2017, 14:11:04) 
[GCC 6.3.0 20170118] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import pandas as pd
>>> import numpy as np
>>> s1 = pd.Series(np.random.randn(5), index=list(range(0, 10, 2)))
>>> s1
0    0.611873
2    1.709006
4   -0.191763
6    1.393935
8   -0.751302
dtype: float64
>>> s1.iloc[:3]
0    0.611873
2    1.709006
4   -0.191763
dtype: float64
>>> s1.iloc[3]
1.3939352812007684
>>> s1.iloc[:3] = 0
>>> s1
0    0.000000
2    0.000000
4    0.000000
6    1.393935
8   -0.751302
dtype: float64


#### With DataFrame
>>> df = pd.DataFrame(np.random.randn(6, 4), index=list(range(0, 12, 2)), columns=list(range(0, 8, 2)))
>>> df
           0         2         4         6
0   0.544588 -0.943141  0.656372 -0.468110
2  -2.093147 -0.122080  0.218309  0.172868
4  -0.529067 -0.418202  1.217741 -0.959730
6  -0.388783 -0.493095 -0.222721  1.224777
8   1.110017  0.786962  1.434934 -1.052975
10  1.945774  0.274088 -0.724777 -0.469002
>>> df.iloc[:3]
          0         2         4         6
0  0.544588 -0.943141  0.656372 -0.468110
2 -2.093147 -0.122080  0.218309  0.172868
4 -0.529067 -0.418202  1.217741 -0.959730
>>> df.iloc[1:5, 2:4]
          4         6
2  0.218309  0.172868
4  1.217741 -0.959730
6 -0.222721  1.224777
8  1.434934 -1.052975
>>> df.iloc[[1, 2], [0, 1]]
          0         2
2 -2.093147 -0.122080
4 -0.529067 -0.418202
>>> df.iloc[:, :3]
           0         2         4
0   0.544588 -0.943141  0.656372
2  -2.093147 -0.122080  0.218309
4  -0.529067 -0.418202  1.217741
6  -0.388783 -0.493095 -0.222721
8   1.110017  0.786962  1.434934
10  1.945774  0.274088 -0.724777
>>> df.iloc[1, 1]
-0.12207998453906502
>>> df.iloc[1]
0   -2.093147
2   -0.122080
4    0.218309
6    0.172868
Name: 2, dtype: float64

mean

计算均值

1
2
3
4
5
6
7
8
9
10
import pandas as pd

# 加载数据
train_df = pd.read_csv('./data/train.csv')

# 计算所有数值均值
train_df.mean()

# 计算某一列均值
train_df['Age'].mean()

dropna

1
DataFrame.dropna(axis=0, how='any', thresh=None, subset=None, inplace=False)
  • axis
    • 0 : 行
    • 1 : 列
    • 元组/列表 : 处理多维数据
  • how
    • any : 只要有一个值为NA，则删除该标签
    • all : 所有值均为NA，则删除该标签
  • thresh
    • 整数值，默认为空
    • 如果非NA值数量大于该值，则保留该标签
  • subset
    • 序列格式参数(array-like)
    • 只考虑给定标签的值
  • inplace
    • True : 在原始数据集上进行操作，返回None
    • False : 返回新的数据集，保持原始数据集不变

删除列

1
df.dorp('column-name', axis=1)

示例程序

1
2
3
4
5
6
7
8
9
10
11
12
13
import pandas as pd


s = pd.Series([1, 2, 3], index=['a', 'b', 'c'])     # 创建一个序列s
d = pd.DataFrame([[1, 2, 3], [4, 5, 6]], columns=['a', 'b', 'c'])   # 创建一个表
d2 = pd.DataFrame(s)

d.head()    # 预览前5行数据
d.describe()    # 数据基本统计量

# 读取文件
pd.read_excel('data.xls')
pd.read_csv('data.csv', encoding='utf-8')
machine learning►tools►pandas

pandas

安装

1
pip3 install pandas

引入

1
import pandas as pd

配置

1
2
3
4
# 修改print时显示列行列数
pd.set_option('display.max_rows', 100)
pd.set_option('display.max_columns', 100)
pd.set_option('display.width', 1000)

导入导出

读取 csv 文件

1
2
3
df = pd.read_csv('dataset.csv')
df = pandas.read_csv('data.csv', delimiter='\t') # 指定 delimiter
df = pd.read_csv("fn", index_col=None, header=0)

从数组创建

1
2
3
4
5
6
7
>>> import pandas as pd
>>> arr = [[1, 2], [3, 4]]
>>> df = pd.DataFrame(arr, columns=['x', 'y'])
>>> df
   x  y
0  1  2
1  3  4

转换为 numpy 对象

1
2
# 转换为 np array
df.to_numpy()

导出到文件

1
df.to_csv('output.cvs')

数据透视

1
2
3
4
5
6
7
8
9
# 描述数据
data.describe()
data.info()

# 打印前 n 行
df.head(n)

# 打印列
data.columns

数据操作

列操作

读操作

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
# 选取一列
data.<ColumnName>

# 选取多列
data[[<ColumnName>, <ColumnName>, <...>]]
# 过滤列
data.loc[:, df.columns != '<column_name>']

# 通过索引选取
df.iloc[:, 0]	# 第1列

# 对列取 unique
df[<ColumnName>].unique()

# 按类型选取列
df.select_dtypes(include=['float64', 'bool'])
df.select_dtypes(include='number') # 选取所有数字类型的列

转换操作

1
2
3
4
5
6
7
8
9
# 转换所有列
df = df.apply(pd.to_numeric)
df = df.apply(pd.to_numeric, errors='ignore')	 # 忽略异常

# 转换特定列
df[['a', 'b']] = df[['a', 'b']].apply(pd.to_numeric)

# 一列转换为多列
df[['a', 'b', 'c']] = df['label'].apply(lambda x: pd.Series(x.split(',')))

Drop 列

1
2
# drop 有 nan 的列
df.dropna(axis=1, how='any')

行操作

排序

1
df = df.sort_values(by=[<ClomunName>], ascending=False)

过滤

1
2
3
4
5
6
7
8
# loc
df.loc[df['column_name'] == some_value]
df.loc[df['column_name'].isin(some_values)]
df.loc[(df['column_name'] >= A) & (df['column_name'] <= B)]

# 按列值过滤
df[df['部门'].isin(['市场部']) & ~df['省份'].isin(['北京'])] # 部门=市场部 & 省份!=北京
t_df = t_df[t_df.服务名称 == '离线计算平台']

连接

1
2
3
4
5
6
7
# 用法
merge(left, right, how='inner', on=None, left_on=None, right_on=None,
      left_index=False, right_index=False, sort=True,
      suffixes=('_x', '_y'), copy=True, indicator=False)

# 左连接
merge(data_x, data_y, how='left', on='uid')

示例

1
2
# 读取 Cell
data['age'][6]  # 读取 data 的 age 列第 6 行数据