Jupyter Notebook使用Anaconda虚拟环境

第一步：安装ipykernel：法一： 1.进入虚拟环境

Windows：在Anaconda Prompt, 运行 activate myenv macOS & Linux, 在Terminal, 运行 source activate myenv 安装ipykernel：conda install ipykernel 法二：在任何环境下都可以直接运行conda install -n myenv ipykernel为myenv安装ipykernel

第二步：将环境写入notebook的kernel中首先要保证在当前虚拟环境下，进入虚拟环境的过程如第一步的法一。运行python -m ipykernel install –user –name=myenv 将环境写入notebook的kernel中

生成配置文件:

1

jupyter notebook --generate-config

生成密码

输入ipython后

1
2
3


from notebook.auth import passwd

passwd()

修改配置文件

1
2
3
4
5
6
7


c.NotebookApp.ip='*' 
c.NotebookApp.notebook_dir = '/home/ozh/share'#共享目录
c.NotebookApp.password = u'sha1:5df252f58b7f:bf65d53125bb36c085162b3780377f66d73972d1' #填写刚刚生成的密文  
c.NotebookApp.open_browser = False # 禁止notebook启动时自动打开浏览器(在linux服务器一般都是ssh命令行访问，没有图形界面的。所以，启动也没啥用)  
c.NotebookApp.port =8899 #指定访问的端口，默认是8899  

c.NotebookApp.allow_remote_access = True

启动

1

jupyter notebook --port=8890 --allow-root

如果要输出图形，请安装

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19


pip config set global.index-url https://pypi.tuna.tsinghua.edu.cn/simple

pip/conda install matplotlib

pip install akshare --upgrade

pip install backtrader

pip install tushare

conda install scipy

anaconda search -t conda tensorflow

conda install tensorflow

conda install pytorch

conda install sympy

# backtrader兼容性问题
pip uninstall matplotlib
pip install matplotlib==3.2.2

conda activate py3

nohup jupyter notebook –port=8890 –allow-root > /root/jupyternotebook.log 2>&1 &

conda activate py3
nohup jupyter notebook --port=8890 --allow-root --NotebookApp.terminals_enabled=False > /root/jupyternotebook.log 2>&1 &
# 禁用控制台命令行
--NotebookApp.terminals_enabled=False

xeus-cling 安装

conda create -n xeus-cling python=3.6

conda activate xeus-cling

conda install -c conda-forge xeus-cling

Anaconda3的安装路径下的/envs/xeus-cling/share/jupyter/kernels/下

jupyter kernelspec install PREFIX/share/jupyter/xcpp11 --sys-prefix
jupyter kernelspec install PREFIX/share/jupyter/xcpp14 --sys-prefix
jupyter kernelspec install PREFIX/share/jupyter/xcpp17 --sys-prefix

jupyter kernelspec install /root/anaconda3/envs/xeus-cling/share/jupyter/kernels/xcpp11 --sys-prefix
jupyter kernelspec install /root/anaconda3/envs/xeus-cling/share/jupyter/kernels/xcpp14 --sys-prefix
jupyter kernelspec install /root/anaconda3/envs/xeus-cling/share/jupyter/kernels/xcpp17 --sys-prefix

/root/anaconda3/envs/xeus-cling/share/jupyter/kernels

firewall-cmd --zone=public --add-port=8891/tcp --permanent
firewall-cmd --reload



conda activate xeus-cling
nohup jupyter notebook --port=8891 --allow-root --NotebookApp.terminals_enabled=True > /root/jupyternotebookc.log 2>&1 &

cmake 安装

yum install -y gcc gcc-c++ make automake

https://cmake.org/download/

sh cmake-3.21.1-linux-x86_64.sh –prefix=/usr/local –exclude-subdir

yum install  -y openssl  openssl-devel 
cd finarthur-master
mkdir build
cd build
cmake ..
make

低延时调优文章

https://zhuanlan.zhihu.com/p/59242346

安装 openonload

## 进入相应的文件夹
cd openonload-201811/

## 源代码存放
cd scripts/

## 搭建环境
./onload_build
## 执行安装
./onload_install
#加载 onload
onload_tool reload

## 停止 cpuspeed 服务以避免进入省电模式，降低CPU时钟速度
systemctl stop cpuspeed

## 停止 irqbalance 服务器以防止 OS 在可用的CPU内核之间重新平衡中断
systemctl stop irqbalance

## 停止防火墙辐射器以消除简介消耗
systemctl stop firewalld

## 禁用 interrupt moderation
ethtool -C enp1s0f1 rx-usecs-irq 0 adaptive-rx off

## 启动低延时配置：tuned-adm
tuned-adm list
tuned-adm profile network-latency

## 防止系统进入 CPU 低功耗模式　cstates
## 参考博文: https://williamlfang.github.io/post/2019-12-11-linux-调整-cstate-实现cpu超频/

Jupyter Notebook使用Anaconda虚拟环境