v mynameisvinn

WITH
  -- userId, candidateId, attribution (eg click, purchase)
  -- candidateid is listingId https://etsy.slack.com/archives/C01T5STE7DH/p1726850552184059?thread_ts=1726850305.245089&cid=C01T5STE7DH
  attribution_tab AS (
  SELECT
    CAST(userId AS int) AS userId,
    candidateId AS listingId
  FROM
    `etsy-sr-etl-prod.etl_data.search_attribution_hourly`

Updates Ranker

Overview

The Updates Ranker ranks notification types (eg thank you coupons).

Surface

The Updates Ranker will be used in the Deal tab.

Runbook

Generate User and Global Notification Features

Notification features (eg user click rate for various notification types, triggers, and sources) are generated via a Sparkly job. These features are moved to FBv2 via the recs_ranking_updates_features_fbv2 DAG

1. Clone your fork:

git clone [email protected]:YOUR-USERNAME/YOUR-FORKED-REPO.git

2. Add remote from original repository in your forked repository:

cd into/cloned/fork-repo
git remote add upstream git://github.com/ORIGINAL-DEV-USERNAME/REPO-YOU-FORKED-FROM.git
git fetch upstream

	from glob import glob
	import json
	from tqdm import tqdm


	# raw treatment -> gs://etsy-recsys-ml-dev-data-nxsn/user/vtang/updates-ranker-v1/metrics/prod_updates_2
	prod_fp = glob("./prod_updates_2/*")
	prod_updates = {} # key = notification_feed_id, value = notification_type of candidate position 0

	for fp in tqdm(prod_fp):

	import requests
	import json
	import numpy as np

	url = 'http://mmx-recsys-ingress.flop.search-infra.etsy-searchinfra-gke-prod-2.etsycloud.com/apiv2/general/rec_set/dynamic/recs'

	headers = {
	'accept': 'application/json',
	'Content-Type': 'application/json',
	}

	# we know the following graph is nilpontent
	m = np.array([[0, 1, 0, 0, 0, 0],
	[0, 0, 1, 1, 0, 0],
	[0, 0, 0, 0, 0, 0],
	[0, 0, 0, 0, 1, 1],
	[0, 0, 0, 0, 0, 1],
	[0, 0, 1, 0, 0, 0]])

	for k in range(len(m)):
	print(m)

	"""
	graph examples collected from https://www.cs.hmc.edu/~keller/courses/cs60/s98/examples/acyclic/

	in the first example, we have an acylic graph. there exists k such its adjacency
	matrix m^k == 0. because we cant produce a path of arbitrarily high length (eg
	greater than k), the graph does not have a cycle.

	in the second example, we have a cyclic graph. there is no k where its adjacency
	matrix m^k == 0.

	def create_tokens(source):
	"""convert source to tokens.

	convert a string "(first (list 1 (+ 2 3) 9))" into a python list
	['(', 'first', '(', 'list', '1', '(', '+', '2', '3', ')', '9', ')', ')'].

	this conversion allows us to recursively inspect each element in the list.
	"""
	return source.replace('(',' ( ').replace(')',' ) ').split()

	class Vertex(object):
	"""represent vertices in a tree/graph."""
	def __init__(self, name):
	self.name = name
	self.children = {}
	self.status = None

	def add_child(self, child):
	key = child.name
	self.children[key] = child