Skip to content

Linux snippets

A collection of useful Linux (and macOS) command-line snippets for data science workflows. All commands are compatible with zsh and bash shells.

List files by size (descending)

1
ls -lhS

Find files larger than 100MB

1
find . -type f -size +100M

Count lines in all CSV files in a directory

1
wc -l *.csv

Show top 10 memory-consuming processes

1
ps aux --sort=-%mem | head -n 11

Search for a pattern in all Python files

1
grep -rnw . -e 'pattern' --include=*.py

Replace text in multiple files (in-place)

1
sed -i '' 's/oldtext/newtext/g' *.txt

Download a file from the internet

1
curl -O https://example.com/file.csv

Extract a tar.gz archive

1
tar -xzvf archive.tar.gz

Monitor disk usage in current directory

1
du -sh *

Show GPU usage (NVIDIA)

1
nvidia-smi

Sample output:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Wed Jun 11 12:18:15 2025       
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.183.01             Driver Version: 535.183.01   CUDA Version: 12.2     |
|-----------------------------------------+----------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |         Memory-Usage | GPU-Util  Compute M. |
|                                         |                      |               MIG M. |
|=========================================+======================+======================|
|   0  NVIDIA H100 PCIe               Off | 00000000:00:08.0 Off |                    0 |
| N/A   37C    P0             117W / 350W |  27085MiB / 81559MiB |     51%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   1  NVIDIA H100 PCIe               Off | 00000000:00:09.0 Off |                    0 |
| N/A   38C    P0              53W / 350W |      3MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+

+---------------------------------------------------------------------------------------+
| Processes:                                                                            |
|  GPU   GI   CI        PID   Type   Process name                            GPU Memory |
|        ID   ID                                                             Usage      |
|=======================================================================================|
|    0   N/A  N/A    213705      C   /usr/bin/python3                          27076MiB |
+---------------------------------------------------------------------------------------+

Check Python package versions in environment

1
pip freeze

Create a virtual environment (Python 3)

1
2
python3 -m venv venv
source venv/bin/activate

Kill a process by name

1
pkill -f process_name

Count unique values in a CSV column

1
cut -d, -f2 file.csv | sort | uniq -c | sort -nr

Preview a CSV file (first 5 rows)

1
head -n 5 file.csv

Check open ports

1
lsof -i -P -n | grep LISTEN

Download all images from a webpage

1
wget -nd -r -P ./images -A jpg,jpeg,png,gif http://example.com

Show the 10 largest files in a directory tree

1
find . -type f -exec du -h {} + | sort -rh | head -n 10

Remove all files except .csv in a directory

1
find . ! -name '*.csv' -type f -delete

Split a large CSV into smaller files (1000 lines each)

1
split -l 1000 bigfile.csv smallfile_

Find the 10 largest directories in the current directory

1
du -h --max-depth=1 | sort -hr | head -n 10

Tmux

Tmux is a terminal multiplexer that allows you to run multiple terminal sessions inside one window. It's especially useful for long-running data science jobs that need to persist after SSH disconnections.

Install tmux (macOS)

1
brew install tmux

Start a new session

1
tmux

Start a new named session

1
tmux new -s my_session

List all sessions

1
tmux ls

Attach to an existing session

1
tmux attach -t my_session

Detach from current session

Press Ctrl+b then d

Kill a session

1
tmux kill-session -t my_session

Split pane horizontally

Press Ctrl+b then "

Split pane vertically

Press Ctrl+b then %

Press Ctrl+b then arrow keys

Create a new window

Press Ctrl+b then c

Switch between windows

Press Ctrl+b then window number (0-9)

Rename current window

Press Ctrl+b then ,

Scroll mode (view history)

Press Ctrl+b then [, use arrow keys to scroll, press q to exit

Run a long training job in tmux

1
2
3
4
tmux new -s training
python train_model.py
# Press Ctrl+b then d to detach
# Reconnect later with: tmux attach -t training

Feel free to copy, modify, and combine these snippets for your data science projects!