common-linux

This is a list of most common Linux command lines tips and tricks. Includes both for C-Shell and Bash.

Linux Admin

Find where commands/files are
```
  locate [keyword]
```
Modifying user
```
  usermod -u uid -g gid [username]
```
IP address
```
  ip a 
```
To go to another terminal screen
```
  ctrl-alt-f[screennumber]  
```

where screen number is the number you want

Kill process (savage)

  kill $(ps aux | grep '[p]ython csp_build.py' | awk '{print $2}')

kill force
```
  kill -9 pid
```
grepping for process
```
  pgrep 'pattern'
```
viewing process tree uncompacted
```
  pstree -c
```
Run program even when ssh connection is terminated
```
  nohup [command] > logfile
```
Find process by a certain user
```
  ps -u username
```
Check logs
```
  dmesg
```
Check cpu usage every 2 seconds for 5 iterations
```
  mpstat 2 5
```
Find process by pattern
```
  ps -fC pattern
```
Kill process by pattern
```
  kill -f pattern
```

ping different hosts

  echo 'set i = 31 \n while( $i < 99 )\n ping -c 2 cpu0$i \n @ i++ \n end' | tcsh | grep '0 received'

Common

Find row based on maximum value of a certain column (looking at second column)
```
  awk 'BEGIN{max=0}{if(($2)>max) {max=($2);i=($1)}END {print max, i}' 
```

Find and move

  find path_A -name '*AAA*' -exec mv -t path_B {} +

Find files large than certain size
```
  find ./ -size +4G
```
Find files/binaries that was access x mins ago
```
  find ./ -amin -x
```
Two ways of executing on find results (ok asks for confirmation)
```
  find ./ -exec rm {} \;
  find ./ -ok rm {} \;
```
Output starting from 25th line
```
  tail -n +25 file.log
```
Finding characters at the end of the line
```
  grep '0,0$' file
```
Print every 10 lines
```
  sed -n "1~10 p" file 
```
Print every 10 lines and 1 line after each of the 10 lines
```
  sed -n "1~10, +1p" file
```
Only report line numbers for grep
```
  grep -n file | cut -f1 -d
```

grep showing filename for occurence of fine

  find ./ -name "Makefile" -exec grep AVX {} /dev/null  \;

Inline for loop (csh)

  echo 'foreach f ( 1 2 3 )\n echo $f \n end' | tcsh

Inline for loop (bash)

  for i in {1..5}; do COMMAND-HERE; done

Count column

  head -2 english_dataset.csv |tail -1  | sed 's/[^,]//g' | wc -c

Replace substring and save back to file
```
  sed -i 's/original/new/g' file.txt
```
Find pid using pattern
```
  pgrep [-afl] [pattern]
```
Kill pids using pattern (-l includes the whole command for the process)
```
  pkill -lf [pattern]
```
Get hostnames and ips registered
```
  getent hosts
```

SeismicUnix

Creating waveform

  uwaveform tape=ricker1 fpeak=(fdommax) dt=(sample rate in seconds) > wavename.su

Windowing (selecting) traces by its keys

  suwind < input.su key=fldr min=1201 max=1301 > test.su

Splitting su data into multiple files based on keys
```
  susplit key=sx < input.su
```
Wave viewing
```
  suxwigb < wavename.su 
```
Get summary of wave (print all traces' header in summary, 0 values omitted)
```
  surange < wave.su 
```

Converting su files to segy files.

  segyhdrs < wavename.su  

  segywrite tape=wavename.segy < wavename.su

Viewing frequency spectrum of wavelet

 sufft < [sufile] | suamp mode=amp | suxwigb

Perc with tpow (for shot gathers)
```
  sugain tpow < [shotgathers.su]
```
Perc with stacked images (migrated)
```
  sugain epow < [image.su]
```
Filter out frequency from a to b so that it's amplitude goes from c to d
```
  sufilter f=a,b amp=c,d < [suwave]
```

Change value of a header property

  suchw  key1=dt key2=scalel a=0 b=25000 e=0 c=0 d=1 < input_su_name > output_su_name

Converting CSV file to segy

  a2b n1=[time steps] < [csv file] | suaddhead ns=[time steps] | sushw key=dt a=[sample rate in micro seconds] > [output file name (.su)]
  segyhdrs < [output file name (.su)]
  segywrite < [output file name (.su)] tape=[output file name (.segy)]

Other commands

Running MPI

  mpirun -np 3 --machinefile hostfile ./executable argument_files.arg

Run only a single testcase for python unittest
```
  python testMyCase.py MyCase.testItIsHot
```

Disable ssl verification for cloning

  git -c http.sslVerify=false clone [addr]

Adding proxy to git (%5C is backslash, %40 is @)

  git config --global https.proxy https://company%5Cuname:password%40@170.28.2.1:8080

Viewing status of tracked files
```
  git status -uno
```
Create a new branch, with modified files to be commited to new branch
```
  git checkout -b [branch name]
```
Revert to certain commit
```
  git reset --hard [commit_id]
```

Pulling from remote with changes locally

  git stash
  git pull origin [remote branch]
  git stash pop

SCons (Madagascar)

Compile everything within directory recursively
```
  sftour scons lock
```
View plots
```
  scons view
```
Create pdf out of paper.tex (*tex file in the directory)
```
  scons pdf
```
Create pdf to read
```
  scons read
```

Python

Modify numpy arrays by column x and y.

  for ix,iy in np.ndindex(placeholder.shape):
  	[operation]

PBSPro

Select pbs jobs by user
```
  qselect -u [username] 
```
View details of pbs job
```
  qstat -fx [pbs_job_id] 
```
deleting pbs job
```
  qdel [pbs_job_id] 
```
View nodes configured under PBS
```
  pbsnodes -aSj 
```

aimanuslim/notebook