mnarayan mnarayan

Google AI Platform Notebook tips

Google Colab: https://colab.research.google.com/

Free compute! and interesting, fixed cost, 10$/month Google Colab Pro version.
Notebooks are very easy to share, with ability to control permissions (integration in google drive/doc ecosystem).
Instant availability. No requirement other than a regular google account.
Idle time limit : 1h in free version.

Go to https://web.telegram.org and sign in if you're not already signed in.
On the top-left, click on the hamburger icon to show the dropdown menu, and choose "Contacts" from the list.
Choose "Edit".
Open the Developer Console of your browser (preferably Chrome or Firefox). On Chrome, that is Ctrl + Shift + J.
Run the following JS snippet:

var x = document.getElementsByClassName('contacts_modal_contact')

	# Make sure to add your SSH key to your RunPod account
	echo "export HF_HOME=/workspace/hf_home/" >> ~/.bashrc
	apt-get update
	apt-get upgrade -y
	apt-get install -y sudo git vim ssh net-tools htop curl zip unzip tmux rsync libopenmpi-dev iputils-ping make fzf restic ripgrep wget pandoc poppler-utils pigz bzip2 nano locales
	pip install uv
	# Alternative: curl -LsSf https://astral.sh/uv/install.sh \| sh
	uv pip install --system --compile-bytecode ipykernel kaleido nbformat numpy scipy scikit-learn scikit-image transformers datasets torchvision pandas matplotlib seaborn plotly jaxtyping einops tqdm ruff basedpyright umap-learn ipywidgets virtualenv pytest git+https://github.com/callummcdougall/eindex.git transformer_lens nnsight
	apt-get install -y libnss3 libatk-bridge2.0-0 libcups2 libxcomposite1 libxdamage1 libxfixes3 libxrandr2 libgbm1 libxkbcommon0 libpango-1.0-0 libcairo2 libasound2
	# Optional:

	import numpy as np
	import numba as nb

	import scipy.stats


	@nb.njit(parallel=True)
	def tiecorrect(rankvals):
	"""
	parallelized version of scipy.stats.tiecorrect

	import matplotlib.pyplot as plt
	x=[-2,-1,0,1,2]; y=[4,1,0,1,4]

	# Set DPI >=150 for readability and adjust figsize
	myfig = plt.figure( figsize=(3,1.5), dpi=150 )

	# Pick a nice color, make sure your markers are large and
	# make sure your lines are wide
	myplt = plt.plot( x, y, 'o', color = 'c', markersize=6)
	plt.plot( x, y, color= 'c', linewidth=2)

	from torch.optim.sgd import SGD
	from torch.optim.optimizer import required
	class PGM(SGD):
	def __init__(self, params, proxs, lr=required, momentum=0, dampening=0,
	nesterov=False):
	kwargs = dict(lr=lr, momentum=momentum, dampening=dampening, weight_decay=0, nesterov=nesterov)
	super().__init__(params, **kwargs)

	if len(proxs) != len(self.param_groups):
	raise ValueError("Invalid length of argument proxs: {} instead of {}".format(len(proxs), len(self.param_groups)))

	# install.packages(c("devtools", "tidyverse"))
	# devtools::install_github("malcolmbarrett/ggdag")
	library(ggdag)
	library(tidyverse)

	dag <- dagify(D ~ ABU1 + AEU2 + BEU3,
	ABU1 ~ A + B + U1,
	AEU2 ~ A + E + U2,
	BEU3 ~ B + E + U3)

	#!/bin/bash
	#
	#SBATCH -J fmriprep
	#SBATCH --array=1-257%7
	#SBATCH --time=24:00:00
	#SBATCH -n 1
	#SBATCH --cpus-per-task=16
	#SBATCH --mem-per-cpu=4G
	#SBATCH -p russpold,owners,normal

	Collections of scripts harvested mainly from Pete, but also picked up from the forums

	TOC

	Accessing dynamic measurements.groovy - Most annotation measurements are dynamically created when you click on the annotation, and
	are not accessible through the standard getMeasurement function. This is a way around that.

	Affine transformation.groovy - access more accurate measurements for the affine transformation used in the image alignment (m5/m6+)

	Alignment of local cells.groovy - check neighborhood for similarly aligned cells

	# An simple function to turn an xts time series
	# into a ggplot calendar heatmap
	require(tidyverse)

	# The core idea is to transform the data such that one can
	# plot "Value" as a function of "WeekOfMonth" versus "DayOfWeek"
	# and facet this Year versus Month

	xts_heatmap <- function(x){
	data.frame(Date=as.Date(index(x)), x[,1]) %>%