To solve this problem use this command:

gpg --keyserver hkp://keyserver.ubuntu.com:80 --recv 9BDB3D89CE49EC21

which retrieves the key from ubuntu key server. And then this:

gpg --export --armor 9BDB3D89CE49EC21 | sudo apt-key add -

Ubuntu无法找到add-apt-repository问题的解决方法

apt-get install python-software-properties software-properties-common

pull an image

docker pull tensorflow/tensorflow:latest-gpu-py3
// docker run -it --rm --runtime=nvidia tensorflow/tensorflow:latest-gpu-py3 python

run a container

docker run -it --rm --runtime=nvidia tensorflow/tensorflow:latest-gpu-py3 python
docker run --gpus all -it -v /home/jinsc/workspace/:/share -v /home/guolab/WIN_Z/jinscZ:/jinscZ --name tfgpu --ipc=host --network=host -p 6006:6006 -p 8888:8888 tensorflow/tensorflow:latest-gpu-py3 /bin/bash

下载所需ubuntu cuda cudnn的docker镜像

http://mccipc.ustc.edu.cn/mediawiki/index.php/Docker-images

pytorch tutorial on ubuntu 16.04 安装pytorch1.1

https://kezunlin.me/post/54e7a3d8/

refereing to: bbest/terrapulse#5

This is the future!

SpaceKnow.com

Deep Learning with Satellite Data

landslide
https://towardsdatascience.com/deep-learning-with-satellite-data-b78b20708de
https://pmm.nasa.gov/landslides/projects.html

Install

refer to the openalea wiki and the doc

https://github.com/openalea/openalea/wiki/install
https://openalea.readthedocs.io/en/latest/install.html

conda create -n openalea python=2.7
activate openalea
conda install -y -c openalea openalea.sconsx openalea.visualea openalea.components openalea.oalab

conda install -c openalea/label/unstable openalea.plantgl
conda install -y -c openalea openalea.mtg openalea.lpy pyqglviewer

PlantGL

https://anaconda.org/openalea/openalea.plantgl

在Anaconda下搜索package，各种conda安装方法尝试，最后ustable的成功了

To install plantgl, I successfully used the command as bellow,
conda install -c openalea/label/unstable openalea.plantgl

visualea

https://openalea.readthedocs.io/en/latest/tutorials/visualea/beginner.html
conda create -n visualea_tuto
activate visualea_tuto
conda install -y -c openalea openalea.visualea openalea.components openalea.plantgl boost=1.66 -c openalea/label/unstable

R_USER not defined

上述设置完成之后Pycharm需要重启，否则pycham中可能还有问题
如果还有问题，参考https://blog.csdn.net/bhcgdh/article/details/81357204
https://blog.csdn.net/lq_520/article/details/83856429?fps=1&locationNum=2

查看版本号

cat /etc/issue

新建环境变量

https://blog.csdn.net/Bleachswh/article/details/51334661
vim /etc/profile
source /etc/profile

更换下载源

更换下载源（中科大）：https://www.linuxidc.com/Linux/2017-01/139319.htm
更换下载源（阿里）：https://www.cnblogs.com/fhychzu/p/7542754.html

批量解压

https://blog.csdn.net/silentwolfyh/article/details/53909518
for i in $(ls *.tar);do tar xvf $i;done
for i in $(ls *.7z);do 7z x $i;done

统计

某文件夹下文件的个数
ls -l |grep "^-"|wc -l

统计某文件夹下目录的个数
ls -l |grep "^ｄ"|wc -l
http://blog.sina.com.cn/s/blog_464f6dba01012vwv.html

docker commit -m="提交的描述信息" -a="作者" 容器id 要创建的目标镜像名:[标签名]

推送到dockerhub

https://blog.csdn.net/boonya/article/details/74906927

1. docker commit -m="message comment" -a='author' containerid imagename:tag
2. docker login
   jinsc
   FA23
3. docker tag imagename:tag jinsc/imagename:tag
4. docker push jinsc/imagename:tag

pull from docker hub

docker pull jinsc/imagename:tag

保存本地

https://blog.csdn.net/anxpp/article/details/51810776
docker save -o /home/jinsc/dockertar/ubuntu16py36torch12cuda10.tar ubuntu16py36torch12cuda10:20190914
或者
docker export 7691a814370e > ubuntu.tar

从本地载入

docker load < ubuntu16py36torch12cuda10.tar

可以使用 docker import 从容器快照文件中再导入为镜像，例如

$ cat ubuntu.tar | docker import - test/ubuntu:v1.0
$ docker image ls
REPOSITORY TAG IMAGE ID CREATED VIRTUAL SIZE
test/ubuntu v1.0 9d37a6082e97 About a minute ago 171.3 MB

https://yeasy.gitbooks.io/docker_practice/container/import_export.html

DeepLearning_ubuntu16.0.4-server.docx

NVIDIA Driver安装

1. 安装 驱动NVIDIA driver
   参考链接http://www.52nlp.cn/%E6%B7%B1%E5%BA%A6%E5%AD%A6%E4%B9%A0%E4%B8%BB%E6%9C%BA%E7%8E%AF%E5%A2%83%E9%85%8D%E7%BD%AE-ubuntu-16-04-nvidia-gtx-1080-cuda-8
   using nvidia-smi to check whether the driver has been successfully installed
   双系统的驱动参考链接：https://blog.csdn.net/zhangk9509/article/details/79300976
   更换下载源（中科大）：https://www.linuxidc.com/Linux/2017-01/139319.htm
   更换下载源（阿里）：https://www.cnblogs.com/fhychzu/p/7542754.html
   驱动下载地址：
   sudo chmod a+x NVIDIA-Linux-x86_64-430.40.run
   sudo sh NVIDIA-Linux-x86_64-430.40.run
   ctrl+alt+F1 进入tty模式
   sudo service lightdm stop #停止显示服务
   sudo service lightdm start

添加Graphic Drivers PPA：
$sudo add-apt-repository ppa:graphics-drivers/ppa
更新Nvidia驱动程序：
$sudo apt-get update
在System Settings -> Software & Updates -> Additional Drivers中选择适配的驱动，点击Apply Changes。

ＣＵＤＡ安装

2, 安装cuda（含NVIDIA driver）
https://developer.nvidia.com/cuda-90-download-archive?target_os=Linux&target_arch=x86_64&target_distro=Ubuntu&target_version=1604&target_type=runfilelocal

sudo chmod a+x cuda_9.0.176_384.81_linux.run
sudo sh cuda_9.0.176_384.81_linux.run

Please make sure that

• PATH includes /usr/local/cuda-9.0/bin
• LD_LIBRARY_PATH includes /usr/local/cuda-9.0/lib64, or, add /usr/local/cuda-9.0/lib64 to /etc/ld.so.conf and run ldconfig as root

安装完毕后，再声明一下环境变量，并将其写入到 ~/.bashrc 的尾部:
export PATH=/usr/local/cuda-8.0/bin${PATH:+:${PATH}}
export LD_LIBRARY_PATH=/usr/local/cuda-8.0/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}

永久添加环境变量https://blog.csdn.net/Bleachswh/article/details/51334661
vim /etc/profile
export PATH=/usr/local/cuda-9.0/bin:$PATH
export LD_LIBRARY_PATH=/usr/local/cuda-9.0/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}

ＣｕＤＮＮ安装

3，安装CuDNN， 选择tar.gz压缩包安装

tar -zxf cudnn-9.0-linux-x64-v7.6.2.24.tgz
cd cuda
ls
sudo cp lib64/* /usr/local/cuda/lib64/
sudo cp include/cudnn.h /usr/local/cuda/include/
cd /usr/local/cuda/lib64/
ls
sudo ln -sf /usr/local/cuda/lib64/libcudnn.so.7.6.2 /usr/local/cuda/lib64/libcudnn.so.7
sudo ln -sf /usr/local/cuda/lib64/libcudnn.so.7 /usr/local/cuda/lib64/libcudnn.so
sudo ldconfig
sudo reboot

Python安装

4. 安装Python环境install Anaconda
   sudo bash Anaconda3-2019.07-Linux-x86_64.sh
   修改环境变量
   sudo gedit /etc/profile
   source /etc/profile
   conda deactivate
   或者
   export Anaconda=/anaconda3/bin/
   export PATH=$PATH:$Anaconda
   source ~/.bashrc

5. 安装PyTorch环境pip install --user torch， torchvision
   pip install --user torch torchvision

6. 可安装Tensorflow等其他环境

Docker安装

1. 安装docker，
https://yeasy.gitbooks.io/docker_practice/content/install/ubuntu.html

docker与nvidia-docker的区别
https://blog.csdn.net/u013355826/article/details/89633619

4. 安装nvidia-docker
the following is installing the docker and nvidia-docker for linux-based deep learning
https://blog.csdn.net/junxiacaocao/article/details/79471770
https://github.com/NVIDIA/nvidia-docker

Install官网安装方式
Add the package repositories
$ distribution=$(. /etc/os-release;echo $ID$VERSION_ID)
$ curl -s -L https://nvidia.github.io/nvidia-docker/gpgkey | sudo apt-key add -
$ curl -s -L https://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.list | sudo tee /etc/apt/sources.list.d/nvidia-docker.list

$ sudo apt-get update && sudo apt-get install -y nvidia-container-toolkit
$ sudo systemctl restart docker

$ sudo service nvidia-docker start
$ sudo nvidia-docker-plugin

Usage

Test nvidia-smi with the latest official CUDA image

$ docker run --gpus all nvidia/cuda:9.0-base nvidia-smi

Start a GPU enabled container on two GPUs

$ docker run --gpus 2 nvidia/cuda:9.0-base nvidia-smi

Starting a GPU enabled container on specific GPUs

$ docker run --gpus '"device=1,2"' nvidia/cuda:9.0-base nvidia-smi
$ docker run --gpus '"device=UUID-ABCDEF,1'" nvidia/cuda:9.0-base nvidia-smi

Specifying a capability (graphics, compute, ...) for my container

Note this is rarely if ever used this way

$ docker run --gpus all,capabilities=utility nvidia/cuda:9.0-base nvidia-smi

If not: 非官网安装方式
wget -P /tmp https://github.com/NVIDIA/nvidia-docker/releases/download/v1.0.1/nvidia-docker_1.0.1-1_amd64.deb
sudo dpkg -i /tmp/nvidia-docker*.deb && rm /tmp/nvidia-docker
sudo service nvidia-docker start
sudo nvidia-docker-plugin

Rancher 安装

5. 安装rancher，拉取镜像使用
https://www.cnblogs.com/gentleman-c/p/7387856.html

1.1 run a rancher
sudo docker run -d --name rancher-server -p 8080:8080 rancher/server
-d表示在后台 –name 为给定的容器的名字 -p端口映射
docker run -it --rm --name ds -p 8888:8888 jupyter/datascience-notebookuna
-it表示交互式的终端 --rm运行玩删除
登录rancher http://10.11.30.151:8080
设置用户名 jinsc 密码 XXXX

#查看公有ip： 159.226.89.150
curl cip.cc 或者 curl ip.cip.cc 或者 curl http://members.3322.org/dyndns/getip

#添加到rancher的主机

nvidia-docker run -it -p 8888:8888 tensorflow/tensorflow:1.10.0-gpu

http://(e5cfdd71b9e0 or 127.0.0.1):8888/?token=437d92fda454fe9759e65a9c26f155ea6b2a50b41bac4299

docker start containerID # docker attach containerID

#主机出现之后查看其中的容器

1.使用-v 参数在run时候挂载

方法１　docker run -it -v /home/jinsc :/share microsoft/dotnet:latest /bin/bash

方法２　使用Volume在主机和容器之间传输文件https://blog.csdn.net/Leafage_M/article/details/78575205

docker volume create my-vol
docker volume ｉｎｓｐｅｃｔ my-vol

运行容器，测试同步数据

2. 对于已经运行中的容器，

docker ps 获取目标容器的ID或者容器的名称 // 我这里的是容器ID为52261df2fab6
docker inspect -f'{{.Id}}' 容器的ID // 获取容器的ID全名称
docker inspect -f'{{.Id}}' c996838f46c1 //返回一个长的ＩＤ名称

有了这个长长的ID的话，本机和容器之间的文件传输就简单了。
docker cp 本地文件路径 ID全称:容器路径
docker cp /home/jinsc/test.abc c996838f46c1261c4422054e3aeda7e301b7645439d3b850b908223751c68cc4:/workspace

docker cp ID全称:容器文件路径 本地路径
docker cp c996838f46c1261c4422054e3aeda7e301b7645439d3b850b908223751c68cc4:/workspace/fromdocker.file /home/jinsc/workspace/

端口映射

jupyter port is 8888　　无法浏览器访问

nvidia-docker run -it -p 8888:8888 tensorflow/tensorflow:latest-gpu

tensorbord port is 6006　 加上0.0.0.0可以再在宿主机上的浏览器打开，也可以在外部电脑打开，外部打开时需要将0.0.0.0换成宿主机的ｉｐ

nvidia-docker run -p 0.0.0.0:6006:6006 -it tensorflow/tensorflow:latest-gpu bash

--network=host 则可以通过外网访问主机的ｉｐ+映射的端口访问容器内部的ｔｅｎｓｏｒborad/jupyter等

nvidia-docker run -it --name PCNN2 --network=host -p 6006:6006 -p 8888:8888 -v /home/jinsc/workspace/:/share ImageID /bin/bash
此时在宿主机本身可通过下面的界面访问，在其他电脑可通过宿主机ｉｐ　１０．１１．３０．１５１:6007查看

数据卷挂载

方法１　docker pull 下载镜像，run　容器，内部自己安装显卡驱动等

在默认情况下Docker是不能访问任何device的, 为了能在Docker里访问显卡,必须加上--privileged=true的选项:
sudo docker run --privileged=true -i -t -v $PWD:/data centos /bin/bash

方法２，nvidia-docker直接下载带有显卡的镜像，减少自己配置的时间

https://zhuanlan.zhihu.com/p/64493662

nvidia-docker run -it -v /home/jinsc/workspace/:/share --name ubuntu16cuda10torch12 --network=host -p 6006:6006 -p 8888:8888 anibali/pytorch:cuda-10.0 /bin/bash

docker run --gpus all -it -v /home/jinsc/workspace/:/share --name ubuntu16cuda10torch12 --network=host -p 6006:6006 -p 8888:8888 anibali/pytorch:cuda-10.0 /bin/bash

docker run --gpus all -it -v /home/jinsc/workspace/:/share -v /home/guolab/WIN_Z/jinscZ:/jinscZ --name superpointgraph --network=host -p 6006:6006 -p 8888:8888 jinsc/ubuntu16cuda9cudnn7py37torch11:superpointgraph /bin/bash

docker run --gpus all -it -v /home/jinsc/workspace/:/share -v /home/guolab/WIN_Z/jinscZ:/jinscZ --name pytorchgeometric --network=host -p 6006:6006 -p 8888:8888 jinsc/ubuntu16cuda9cudnn7py37torch11:pytorchgeometric /bin/bash

加上--ipc=host 防止出现ERROR: Unexpected bus error encountered in worker. This might be caused by insufficient shared memory (shm).

docker run --gpus all -it -v /home/jinsc/workspace/:/share -v /home/guolab/WIN_Z/jinscZ:/jinscZ --name pytorchgeometric --ipc=host --network=host -p 6006:6006 -p 8888:8888 jinsc/ubuntu16cuda9cudnn7py37torch11:pytorchgeometricVIM /bin/bash

docker run --gpus all -it -v /home/jinsc/workspace/:/share -v /home/guolab/WIN_Z/jinscZ:/jinscZ --name superpointgraph2 --ipc=host --network=host -p 6006:6006 -p 8888:8888 jinsc/ubuntu16cuda9cudnn7py37torch11:superpointgraphVIM /bin/bash

Test nvidia-smi with the latest official CUDA image

$ docker run --gpus all nvidia/cuda:9.0-base nvidia-smi

Start a GPU enabled container on two GPUs

$ docker run --gpus 2 nvidia/cuda:9.0-base nvidia-smi

Starting a GPU enabled container on specific GPUs

$ docker run --gpus '"device=1,2"' nvidia/cuda:9.0-base nvidia-smi
$ docker run --gpus '"device=UUID-ABCDEF,1"' nvidia/cuda:9.0-base nvidia-smi

Specifying a capability (graphics, compute, ...) for my container

Note this is rarely if ever used this way

$ docker run --gpus all,capabilities=utility nvidia/cuda:9.0-base nvidia-smi

Reference

http://briancaffey.github.io/2017/11/20/using-tensorflow-and-tensor-board-with-docker.html

一个很好的学习视频

https://www.youtube.com/watch?v=sB07YOwW0SY

搜索镜像

docker search ubuntu

拉取镜像

docker pull ubuntu:16.04

显示镜像

docker images
docker image ls

启动容器

docker run -it --name ubuntu16py3tf16 ubuntu:16.04
或者 （默认启动bash）
docker run -it --name ubuntu16py3tf16 ubuntu:16.04 bash
或者加一个 --provileged 让docker实现root权限
docker run -it --name ubuntu16py3tf16 --privileged ubuntu:16.04 bash
nvidia-docker run -it -d -v /home/jinsc/workspace/:/root/workspace --name nvidiatf ubuntu:16.04

容器内安装tensorflow

容器内升级一下 apt-get

apt-get update
apt-get install libssl-dev
apt-get install make
apt-get install gcc

装python之前先装wget

apt-get install wget

去网上找python的下载链接

mkdir tools
cd tools
wget https://www.python.org/ftp/python/3.6.0/Python-3.6.0.tgz

解压安装python

tar -zxvf Python-3.6.0.tgz
cd Python-3.6.0
./configure
make
make install

安装tensorflow

apt-get install python-pip python-dev
pip3 install tensorflow-gpu
pip3 install tensorflow-gpu==1.6 指定tensorflow版本

检查安装是否成功

python3
import tensorflow as tf
tf._ version _

辅助性开发工具

apt-get install git
apt-get install vim

容器commit生成镜像

docker commit containerID
docker images 查看镜像

镜像保存本地

docker save 镜像名 > /tmp/givename.tar

从本地加载镜像

docker load < /tmp/givename.tar

显示容器

docker container ls
docker container ls -a

docker ps
docker ps -a

进入已经退出的容器

docker start containerID
docker attach containerID

docker exec containerID bash

正常退出不关闭容器，请按Ctrl+P+Q进行退出容器

容器与宿主机之间拷贝文件

docker cp nvidiatf:/workspace/hello.py /home/jinsc/workspace/
docker cp 容器名：/容器内路径/文件 /宿主机目标路径/

删除容器和镜像

docker image rm XXX
docker container stop XX docker container rm XXX
https://blog.csdn.net/CSDN_duomaomao/article/details/78587103

启动并配置配置rancher

sudo docker run -d --name rancher-server -p 8080:8080 rancher/server
查看共有ip
curl cip.cc 或者 curl ip.cip.cc 或者 curl http://members.3322.org/dyndns/getip

tensorflow docker 使用

https://www.tensorflow.org/install/docker?hl=zh-cn

docker run -it -d -rm --name XX -v ~/hostdir:/notebooks -p 8888:8888 -p 6006:6006 tensorflow/tesnsorflow
docker exec -it tf tensorborad --logdir tf_logs/
参数的意义：
-it 交互式终端
-rm 容器停止后就自动删除
-d 参数表示后台守护进程运行容器
--name 参数表示容器的名称，可随意取
-v 表示主机和容器共享文件的映射，容器的目录就是 Dockerfile 中用 VOLUME 命令定义的目录
-p 表器主机和容器端口映射，容器的端口就是 Dockerfile 中用 EXPOSE 命令绑定的端口

进入容器的bash命令行

docker exec -it 83fd3bb36025 bash
相比于 docker attach containerID 进入默认的命令， docker exec更灵活

1.下载window服务器的镜像ios， 利用UltraISO 或者 rufus 软件制作启动优盘；插入到服务器，重启过程按键盘上有点的那个键del，或者F2，F11或者其他的键（根据自己的电脑牌子主板查一下）进入开机启动界面。进入之后更换开机启动项为从优盘启动，启动之后进入重装系统界面。
windows2016server64bit.iso 链接：https://pan.baidu.com/s/1yenJ71THEwShPxxLYWzlpw 提取码：1m7l
windows2019server64bit.iso 链接：https://pan.baidu.com/s/1O0KOSF8GUAdoZY_fTXnRhQ 提取码：nqm3
windows2016--2019镜像说明： 链接：https://pan.baidu.com/s/1SljjynVfqb5x1TkH35uVRw 提取码：5scb

**2. 安装nvidia 驱动，**下列链接选择对应平台的驱动【安装过程电脑会重启几次，提前保存其他程序文件】
https://www.nvidia.com/Download/index.aspx

#**3. 安装cuda,**下列链接选择对应平台的驱动
https://developer.nvidia.com/cuda-downloads?target_os=Windows&target_arch=x86_64

#--------------Cudnn可选安装，
CuDNN 下载地址，需要注册登录，可以用微信等登录；https://developer.nvidia.com/rdp/form/cudnn-download-survey

#-------visualstudio2015（c++）版本也可选安装

4. 安装Anoconda3 版本的python https://www.anaconda.com/distribution/， 安装目录可以选择在D：\anaconda3, 放在某一个盘的根目录下比较方便；

#5. 在pytorch官网选择对应的平台，利用pip install torch 和torchvision 【截图中两行命令，pip3 用pip； pip的位置是在D:\anaconda3\scripts\pip.exe】

安装完成之后，在python中运行，
import torch
print(torch.cuda.is_available()) #如果返回True，说明安装GPU版本的torch成功了

#6.配置editplus,

https://thecustomizewindows.com/2017/07/install-anaconda-ubuntu-16-04/

94 apt update
95 apt upgrade
96 ls
97 cd tmp/
98 ls
99 curl -O https://repo.continuum.io/archive/Anaconda3-4.4.0-Linux-x86_64.sh
100 chmod +x Anaconda3-4.4.0-Linux-x86_64.sh
101 bash Anaconda3-4.4.0-Linux-x86_64.sh
102 source ~/.bashrc

https://blog.csdn.net/ydyang1126/article/details/77247654

import matplotlib as mpl
mpl.use('Agg')
from matplotlib import pyplot as plt
import numpy as np
x = np.linspace(0, 3, 100)
y=x
plt.plot(x,y)
[<matplotlib.lines.Line2D object at 0x7fadda5e8358>]
plt.savefig('3.png')

安装http://fisadev.github.io/fisa-vim-config/

apt update
apt install vim
apt install curl vim exuberant-ctags git ack-grep
pip install pep8 flake8 pyflakes isort yapf

git clone https://github.com/fisadev/fisa-vim-config
cd fisa-vim-config/ # 隐藏文件，ls -a
cp .vimrc ~/.vimrc

加上F5执行命令 https://stackoverflow.com/questions/18948491/running-python-code-in-vim
编辑查看复制
map :exec '!python' shellescape(@%, 1)

插件管理https://blog.csdn.net/zhangpower1993/article/details/52184581

安装需要的插件
将想要安装的插件，按照地址填写方法，将地址填写在vundle#begin和vundle#end之间就可以
保存之后，有两种方法安装插件。
(1) 运行 vim ,再运行 :PluginInstall
(2) 通过命令行直接安装 vim +PluginInstall +qall
移除不需要的插件
编辑.vimrc文件移除的你要移除的插件所对应的plugin那一行。
保存退出当前的vim
重新打开vim，输入命令BundleClean。
其他常用命令
更新插件BundleUpdate
列出所有插件BundleList
查找插件BundleSearch

try

touch a.py
vim a.py

using pdb for python debug

https://blog.csdn.net/gwzz1228/article/details/78059651

cmake error in ply_c

查找“cmake clean”命令以清除cmake输出
时间: 2017-03-29 08:13:49.0标签: cmake
译文: 来源翻译纠错
正如make clean删除makefile生成的所有文件，我想对CMake做同样的。我经常发现自己手动通过目录删除像cmake_install.cmake和CMakeCache.txt文件和CMakeFiles文件夹。
rm -rf CMakeCache.txt cmake_install.cmake CMakeFiles/

CONDAENV=/root/anaconda3 #YOUR_CONDA_ENVIRONMENT_LOCATION
cd partition/ply_c
cmake . -DPYTHON_LIBRARY=$CONDAENV/lib/libpython3.7m.so -DPYTHON_INCLUDE_DIR=$CONDAENV/include/python3.7m -DBOOST_INCLUDEDIR=$CONDAENV/include -DEIGEN3_INCLUDE_DIR=$CONDAENV/include/eigen3
make
cd ..
cd cut-pursuit
mkdir build
cd build
cmake .. -DPYTHON_LIBRARY=$CONDAENV/lib/libpython3.7m.so -DPYTHON_INCLUDE_DIR=$CONDAENV/include/python3.7m -DBOOST_INCLUDEDIR=$CONDAENV/include -DEIGEN3_INCLUDE_DIR=$CONDAENV/include/eigen3
make

run semantic3D

SEMA3D_DIR=/jinscZ/superpoint_graph/learning/datasets/semantic3d

python partition/partition.py --dataset sema3d --ROOT_PATH $SEMA3D_DIR --voxel_width 0.05 --reg_strength 0.8 --ver_batch 5000000

python learning/sema3d_dataset.py --SEMA3D_PATH $SEMA3D_DIR

官方文档

https://jupyterlab.readthedocs.io/en/stable/user/extensions.html

打开方式

(base) > jupyter-lab
然后浏览器会弹出来

插件安装

jupyter labextension install @jupyter-widgets/jupyterlab-manager
jupyter labextension install jupyter-matplotlib

从github clone的项目，在文件夹内增加文件

https://www.jianshu.com/p/f7d999f9ecce

在本地文件夹init新的repository，推送到github

快捷键

Pycharm下面出现No R interpreter defined

https://www.cnblogs.com/qianzf/p/10620407.html

Python在终端通过pip安装好包以后，在Pycharm中依然无法使用的解决办法

https://blog.csdn.net/kouyi5627/article/details/80531442

增加

conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/noarch
conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/win-64
conda config --add channels https://conda.anaconda.org/openalea/noarch
conda config --add channels https://conda.anaconda.org/openalea/win-64

删除

conda config --remove channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/noarch

查看源

conda config --show-sources

查看所有配置信息

conda info

登录登出

anaconda login 用户名jinsc 密码F--A3
anaconda logout

##conda 创建虚拟环境
conda create -n envname python=3.6

conda 装包

conda install pyntcloud -c conda-forge

Still trying ------

method1

https://www.douban.com/note/637559808/

other optional links

https://blog.csdn.net/Totoro1745/article/details/75062729

conda list

conda install python=3.6

conda install pytorch=0.4.1 cuda90 -c pytorch
https://pytorch.org/get-started/previous-versions/

方法１　docker pull下载镜像，run容器，内部自己安装显卡驱动等

在默认情况下Docker是不能访问任何device的, 为了能在Docker里访问显卡,必须加上--privileged=true的选项:
sudo docker run --privileged=true -i -t -v $PWD:/data centos /bin/bash

方法２，nvidia-docker直接下载带有显卡的镜像，减少自己配置的时间

https://zhuanlan.zhihu.com/p/64493662

工作目录设置