Useful unix commands

Posted on | In
| Words count in article 589

https://stackoverflow.com/questions/43702546/tensorboard-doesnt-show-all-data-points

1
tensorboard --samples_per_plugin scalars=0 --logdir xxx/

For Mac OS, the caffeinate command is used to prevent a Mac from going to sleep.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
# stay until you tell the command to stop running 
# or close the terminal
caffeinate 

# wait for the process with ID 41734 to finish
caffeinate -w 41734  

# 18000 is the timeout value in seconds 
caffeinate -t 18000  

# Option -d prevents the display from sleeping.
# Option -i prevents the system from idle sleeping.
# Option -s prevents the system from sleeping. 
# Option -s is valid only when system is running on AC power.
# Some other options also exist.

To get the summary of a grand total disk usage size of a directory in “Human Readable Format”:

1
2
du -sh *
# 1.8G	XXX

Report file system disk space usage. For example, to get the amount of the remaining disk size in “Human Readable Format”:

1
2
3
4
5
# information of the system
df -h

# information of a directory
df -h *

On Mac OS, check CPU temperature (this needs to be installed sudo gem install istats)

1
istats

On Mac OS, change the default shell to Zsh

1
chsh -s /bin/zsh

On Mac OS, change the default shell to Bash

1
chsh -s /bin/bash

Check sum on Mac OS

1
2
3
4
5
# user manual
shasum --help

# check sha1 sum 
shasum -a 1 [file_name]

zip using command line

1
2
3
4
5
6
7
8
9
10
11
# zip a directory
zip -r new_zip_file.zip target_directory/

# zip all the content in the current directory 
zip -r new_zip_file.zip . 

# -X excludes extra file attributes
zip -r -X new_zip_file.zip target_directory/

# zip several files 
zip new_zip_file.zip target_file_1 target_file_2

unzip using command line

1
2
3
4
5
6
# unzip into a separate new directory 
unzip -d name_of_a_new_directory blabla.zip

# unzip and put all the files directly 
# into the current directory
unzip blabla.zip

NVIDIA System Management Interface (nvidia-smi)

1
2
3
4
5
6
7
8
nvidia-smi

# If we just want the GPU name
nvidia-smi --query-gpu=name --format=csv,noheader

# The following will loop and call 
# the view at every second. 
nvidia-smi -l 1

Print newline, word, and byte counts for each file

1
2
3
4
5
6
7
wc

# count the number of lines
wc -l

# count the number of files in [name_of_directory]
ls [name_of_directory] | wc -l

1
2
3
4
5
6
7
8
9
10
11
12
13
man wc

-c, --bytes
    print the byte counts

-m, --chars
    print the character counts

-l, --lines
    print the newline counts
    
-w, --words
    print the word counts

Search pattern (regular expression)

1
2
3
4
grep

# print the line containing the pattern [pattern]
ls [name_of_directory] | grep [pattern]

Search the occurences of some text in all files in a folder.

1
grep -rl "blabla" [folder_path]

grep 是一个在文件中搜索文本的命令。 -r 选项表示递归搜索子目录。 -l 选项表示只列出包含匹配文本的文件名。 "blabla" 是要搜索的特定单词。 [folder_path] 是你要搜索的目录的路径。

如果你要在当前目录及其子目录中搜索,

1
grep -rl "blabla" .

--include="*.py" 选项表示只搜索扩展名为 .py 的文件。

1
grep -rl --include="*.py" "blabla" [folder_path]