Enes Ergin eneskemalergin

"Why Should I Trust You?": Explaining the Predictions of Any Classifier

The paper introduces a novel technique to explain the predictions of any classifier in an interpretable and faithful manner.
It also proposes a method to explain models by obtaining representative individual predictions and their explanations.
Link to the paper
Demo

A personal diary of DataFrame munging over the years.

Convert Series datatype to numeric (will error if column has non-numeric values)
(h/t @makmanalp)

sudo apt-get install python-setuptools

	import numpy as np

	EPSILON = 1e-10


	def _error(actual: np.ndarray, predicted: np.ndarray):
	""" Simple error """
	return actual - predicted

	"""
	Minimal character-level Vanilla RNN model. Written by Andrej Karpathy (@karpathy)
	BSD License
	"""
	import numpy as np

	# data I/O
	data = open('input.txt', 'r').read() # should be simple plain text file
	chars = list(set(data))
	data_size, vocab_size = len(data), len(chars)

	#!/usr/bin/env bash
	# Ubuntu Server or VM Cleaner. Safe by default; aggressive when asked.
	# Example safe: sudo ./clean.sh
	# Example aggressive: sudo JOURNAL_DAYS=3 AGGRESSIVE=1 ./clean.sh
	# Enable Docker image prune (images only): sudo ./clean.sh --docker-images
	# Tested on Ubuntu 20.04, 22.04, 24.04 (server/VM images).

	set -Eeuo pipefail
	trap 'rc=$?; echo "Error on line $LINENO: $BASH_COMMAND (exit $rc)"; exit $rc' ERR
	IFS=$'\n\t'

	The [Cufflinks](http://cole-trapnell-lab.github.io/cufflinks/) Unix package is extremely well regarded for RNA-seq analysis. Yet it is widely acknowledged that installation is difficult, not least by the [Cufflinks authors themselves](http://cole-trapnell-lab.github.io/cufflinks/getting_started/#installing-a-pre-compiled-binary-release)! Whilst the authors provide precompiled binaries to make the installation process easier [the instructions aren't very explicit](http://cole-trapnell-lab.github.io/cufflinks/getting_started/). Having spent a good hour on Google and StackOverflow I've put together an instruction list for the less experienced Unix user like myself.

	Note this guide is for Unix distributions.

	1. Download the Cufflinks tar from http://cole-trapnell-lab.github.io/cufflinks/install/
	'''shell
	wget http://cole-trapnell-lab.github.io/cufflinks/install/
	'''
	2. Untar the downloaded file
	'''shell

	import numpy as np
	from scipy import stats
	from itertools import combinations
	from statsmodels.stats.multitest import multipletests
	from statsmodels.stats.libqsturng import psturng
	import warnings


	def kw_dunn(groups, to_compare=None, alpha=0.05, method='bonf'):
	"""

	#!/usr/bin/env sh

	# first check to see if mongo service is running. you can't delete any files until the service stops so perform a quick check.
	launchctl list \| grep mongo
	# NOTE: the pipe \| symbol means the commands on the right performs on the output from the left
	# grep is a string search utility. `grep mongo` means search for the substring mongo

	# use the unload command to end the mongo service. this is required to 'unlock' before removing the service.
	# first look for the file to delete
	MONGO_SERVICE_FILE=$(ls ~/Library/LaunchAgents/mongodb)