ubuntu 16.04 安装anaconda tensorflow opencv keras openslide-python pycharm

╰+攻爆jí腚メ 2022-03-27 11:12 244阅读 0赞

# 安装anaconda #

首先在anaconda的清华镜像 [https://mirrors.tuna.tsinghua.edu.cn/anaconda/archive/][https_mirrors.tuna.tsinghua.edu.cn_anaconda_archive]

下载anaconda3-5.1.0-Linux-x86-64.sh

$ sudo bash Anaconda3-5.1.0-Linux-x86\_64.sh

然后Do you accept the license terms? \[yes|no\]  
输入yes

我直接安装在默认目录：Anaconda3 will now be installed into this location:  
/home/cassiel/anaconda3

Do you wish the installer to prepend the Anaconda3 install location  
to PATH in your /home/cassiel/.bashrc ? \[yes|no\]  
\[no\] >>> yes  
就会自动加入环境变量

$ sudo vim ~/.bashrc

可以看到末尾如图：

![2019011313323610.png][]

加入

$ export PATH="$PATH:$HOME/bin"

$ alias pyana3="/home/cassiel/anaconda3/bin/python3"    （写自己的用户名）

![20190113134052662.png][]

然后 输入 $ source ~/.bashrc 更新一下

# 安装tensorflow #

首先创建虚拟环境：

conda create -n tf

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl8zNzQ4NTcwOA_size_16_color_FFFFFF_t_70][]

激活虚拟环境：（就是进入到这个虚拟环境里面）

source activate tf

PS:如果要退出：输入 source deactivate tf    移除某环境 conda remove --n 环境名 --all

更改anconda下载镜像为清华仓库镜像（这样舒速度快一点）  
打开Anaconda Prompt，输入清华仓库镜像，这样更新会快一些：

conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/  
conda config --set show\_channel\_urls yes

激活虚拟环境（进入虚拟环境）后

输入 conda install tensorflow

然后等着吧。

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl8zNzQ4NTcwOA_size_16_color_FFFFFF_t_70 1][]

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl8zNzQ4NTcwOA_size_16_color_FFFFFF_t_70 2][]

再次进入python3.6,可以导入tensorflow了，（第一次打错tensorflow，吓死了）

退出镜像（source deactivate tf）在终端中输入 anaconda-navigator 可以打开图像管理界面

可以看到tf镜像和root  
![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl8zNzQ4NTcwOA_size_16_color_FFFFFF_t_70 3][]

# 启动Tensorflow开发环境 #

有两种方法，

方法一，通过anaconda启动：

anaconda-navigator  
\#启动anaconda->home->(Applications on)tensorflow，启动相关开发软件

方法二，通过terminal启动：

source activate tensorflow \# 激活Tensorflow环境  
python \# 或ipython或Spyder启动开发环境

# 安装opencv #

参考： [https://blog.csdn.net/isuccess88/article/details/73546835][https_blog.csdn.net_isuccess88_article_details_73546835]

在这个里面下载 [https://anaconda.org/menpo/opencv3/files][https_anaconda.org_menpo_opencv3_files] 我下载的是 opencv3-3.1.0-py36\_0.tar.bz2

进入tf虚拟环境 source activate tf

conda install opencv3-3.1.0-py36\_0.tar.bz2

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl8zNzQ4NTcwOA_size_16_color_FFFFFF_t_70 4][]

可以导入opencv2

# 安装keras #

在虚拟环境输入 pip install keras

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl8zNzQ4NTcwOA_size_16_color_FFFFFF_t_70 5][]

如果遇到 pip install 提示权限不够 ，可以添加环境变量  export PATH=~/.local/bin:$PATH

参考至： [https://blog.csdn.net/tanzuozhev/article/details/77585342][https_blog.csdn.net_tanzuozhev_article_details_77585342]

我习惯用jupyter，同样在虚拟环境输入conda install jupyter即可

再打开jupyter notebook 就可以了！

# 安装openslide-python #

sudo apt-get install openslide-tools

sudo apt-get install python-openslide

进入虚拟环境

pip install openslide-python

可以import openslide

# 安装pycharm #

[https://www.jetbrains.com/pycharm/download/previous.html][https_www.jetbrains.com_pycharm_download_previous.html]

找到linux版本，我用社区版足以

解压在 /文档 下，然后移动到 /usr/local下

sudo mv pycharm-community-2018.3.2/ /usr/local

cd /usr/local/pycharm-community-2018.3.2/bin

./pycharm.sh

安装路径 /usr/local/bin/charm

安装成功后，打开file-setting，选择project interpreter，小齿轮，选择add

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl8zNzQ4NTcwOA_size_16_color_FFFFFF_t_70 6][]

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl8zNzQ4NTcwOA_size_16_color_FFFFFF_t_70 7][]

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl8zNzQ4NTcwOA_size_16_color_FFFFFF_t_70 8][]

参考：

[https://blog.csdn.net/arpospf/article/details/80228587][https_blog.csdn.net_arpospf_article_details_80228587]

[https://blog.csdn.net/m0\_37864814/article/details/82112029][https_blog.csdn.net_m0_37864814_article_details_82112029]

[https://blog.csdn.net/qq\_28888837/article/details/80757731][https_blog.csdn.net_qq_28888837_article_details_80757731]

[https://blog.csdn.net/hgdwdtt/article/details/78633232][https_blog.csdn.net_hgdwdtt_article_details_78633232]

pycharm 参考文档： [https://blog.csdn.net/johinieli/article/details/79575169][https_blog.csdn.net_johinieli_article_details_79575169]

[https://blog.csdn.net/liangyihuai/article/details/77842628][https_blog.csdn.net_liangyihuai_article_details_77842628]

[https://blog.csdn.net/reallocing1/article/details/53375042][https_blog.csdn.net_reallocing1_article_details_53375042]

[https://blog.csdn.net/zhuanshu666/article/details/73554885][https_blog.csdn.net_zhuanshu666_article_details_73554885]

[https_mirrors.tuna.tsinghua.edu.cn_anaconda_archive]: https://mirrors.tuna.tsinghua.edu.cn/anaconda/archive/
[2019011313323610.png]: /images/20220327/4f058800a7384dc5959479ae1330a163.png
[20190113134052662.png]: /images/20220327/835eebdddbf84fd2bb456936ff0fed99.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl8zNzQ4NTcwOA_size_16_color_FFFFFF_t_70]: /images/20220327/8ac7b12b80ed49218b1220c6ee8441d1.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl8zNzQ4NTcwOA_size_16_color_FFFFFF_t_70 1]: /images/20220327/0f7cfe00552548d7918320878ac8a60c.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl8zNzQ4NTcwOA_size_16_color_FFFFFF_t_70 2]: /images/20220327/f89171f9c2d7454998c9b4629906196f.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl8zNzQ4NTcwOA_size_16_color_FFFFFF_t_70 3]: /images/20220327/07b1dc14c1064ff9abf39e0445a1fb64.png
[https_blog.csdn.net_isuccess88_article_details_73546835]: https://blog.csdn.net/isuccess88/article/details/73546835
[https_anaconda.org_menpo_opencv3_files]: https://anaconda.org/menpo/opencv3/files
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl8zNzQ4NTcwOA_size_16_color_FFFFFF_t_70 4]: /images/20220327/4bffc436ae85443e8157da14746eb394.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl8zNzQ4NTcwOA_size_16_color_FFFFFF_t_70 5]: /images/20220327/b29738a070794b08bad9a71f3c67323c.png
[https_blog.csdn.net_tanzuozhev_article_details_77585342]: https://blog.csdn.net/tanzuozhev/article/details/77585342
[https_www.jetbrains.com_pycharm_download_previous.html]: https://www.jetbrains.com/pycharm/download/previous.html
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl8zNzQ4NTcwOA_size_16_color_FFFFFF_t_70 6]: /images/20220327/acf182429d4349429d85e2c356405b10.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl8zNzQ4NTcwOA_size_16_color_FFFFFF_t_70 7]: /images/20220327/f97c34d16c064da9bb323863e571f386.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl8zNzQ4NTcwOA_size_16_color_FFFFFF_t_70 8]: /images/20220327/d3fbe948ff0d4e7eb9eff2963d0b4cbd.png
[https_blog.csdn.net_arpospf_article_details_80228587]: https://blog.csdn.net/arpospf/article/details/80228587
[https_blog.csdn.net_m0_37864814_article_details_82112029]: https://blog.csdn.net/m0_37864814/article/details/82112029
[https_blog.csdn.net_qq_28888837_article_details_80757731]: https://blog.csdn.net/qq_28888837/article/details/80757731
[https_blog.csdn.net_hgdwdtt_article_details_78633232]: https://blog.csdn.net/hgdwdtt/article/details/78633232
[https_blog.csdn.net_johinieli_article_details_79575169]: https://blog.csdn.net/johinieli/article/details/79575169
[https_blog.csdn.net_liangyihuai_article_details_77842628]: https://blog.csdn.net/liangyihuai/article/details/77842628
[https_blog.csdn.net_reallocing1_article_details_53375042]: https://blog.csdn.net/reallocing1/article/details/53375042
[https_blog.csdn.net_zhuanshu666_article_details_73554885]: https://blog.csdn.net/zhuanshu666/article/details/73554885