[1] Compress PDF (Source: Stack Overflow)

1a using postscript as Intermediate step

Step 1:Convert to intermediate postscript format

pdf2ps large.pdf large.ps

Step2 Convert ps back to pdf

ps2pdf large.ps small.pdf

1b Using ghostscript

gs -dQUIET -dBATCH -dNOPAUSE -sDEVICE=pdfwrite -dPDFSETTINGS=/printer -sOutputFile=output.pdf Sarvesh_BOB_filled_scanned_final.pdf

[2] Run a Bash command every 10 seconds

Syntax

watch -n10 "args command"

Example - Command to list files in the current directory

Example: watch -n10 "ls -ltr | wc"

[3] Convert all files in a folder - Example Convert files between jpg and png

for i in *.jpg ; 
  do 
    convert "$i" "${i%.*}.png" ; 
    
    
  done

[4] Set bash prompt

export PS1= "\u@\h: \w:$"

where \u is username, \h is hostname, \w is present working directory

example: export PS1="$(whoami)@$(hostname):$(pwd)"

[5] Search for files with multiple extensions - example: Images with both `.png` and `.jpg` format

find <file_path> -iname "*.jpg" -or -iname "*.png" | wc

[6] Copy files from remote machine to local machine with following symbolic option

rsync -avzL username@remote:path local-path

[7] Changing jupyter notebook themes

Install jupyterthemes and run the following command for monokai font jt -t monokai -cellw 90%

For list of all available themes use jt -l

[8] Checking Webcam available and ready for use

Checking if webcam detected by Ubuntu

• Use this command lsusb to list all devices detected

Take test photos/video using Cheese

• Go to terminal, type cheese and a window should pop-up. You can take phots(single/burst) and record video!
• In case cheese is not installed, you can install it using sudo apt-get install cheese

[9] Download large zip files from dropbox link

• wget <url>
• Check integrity of zip file using unzip -t <zipfile_name>

[10] Read Image and overlay rectangle using Python

from matplotlib import pyplot as plt
from PIL import Image
from matplotlib.patches import Rectangle

image = Image.open(<path_to_img>)
# rectangle co-ordinates is of the form (xmin,ymin), width, height
origin = (xmin, ymin)
width  = xmax-xmin
height = ymax-ymin

# display image and overlay rectangle
plt.imshow(image)
patch = Rectangle(origin, width, height, linewidth = 10, edgecolor='r')
plt.gca().add_patch(patch)
plt.show()

[11] Get Laptop configuration from command line

• sudo lshw -html > config.html to dump specification as html file
• Use firefox config.html to view the details

[12] Git stop tracking a file

git update-index --skip-worktree <file_name>

[13] Saving and recreating a conda virtual environment

# activate conda environment
conda activate <env_name>

# export environment to a file
conda env export  > environment.yml

# recreate conda environment from file
conda env create -f environment.yml

### ```[14] Pytorch enable GPU/CPU based on availability```

```python
device = torch.device("cuda:0" if torch.cuda_is_available() else "cpu")

[15] Pytorch choose GPU based on availability

Option 1:

os.system("export CUDA_VISIBLE_DEVICES = 0,1")

Option 2: Using bash commandline

export CUDA_VISIBLE_DEVICES = 0,1