Data Science

The goal of every search engine marketing (SEM) advertiser is to maximize their returns at the lowest possible cost. Campaign performance is primarily tuned by adjusting the maximum cost per click (CPC) bid for each ad. However, finding the “perfect” CPC bid can be a moving target since the auction is constantly in flux. The “sleeper” problem Imagine an extreme (but likely) scenario where ad spend is significantly over the allocated budget for the month. The SEM expert can do a few things to quickly get the account back in shape: ...

The US has a peculiar relationship with guns where we frequently observe nontrivial spikes in firearm sales. These are triggered (pun intended) by various political, economic, and social events at the time. With 2020 being an especially chaotic year, I wanted to explore how that phenomenon is reflected in American gun purchases thus far. Load modules import pandas as pd import matplotlib.pyplot as plt import seaborn as sns Retrieve data I was unable to find confirmed and accurate gun purchase data that is free on the web. However, we can use data from the FBI’s National Instant Criminal Background Check System (NICS) as a proxy. ...

With mortgage rates at a historical low there are inklings the US housing market is heating up again. Buying a home is a huge decision and in a perfect world everyone weighs their options and makes a (relatively) rational choice. One approach is to lay out all the mortgage offers and compare how much more we’re paying over the life of the loan. In this article I want to achieve a few things: ...

With the end of the latest Star Wars trilogy, I wanted to compare, contrast, and explore Lucas vs Disney era domestic box office revenue. The analysis and python code below will parse weekly ticket sales from Box Office Mojo, adjust revenue numbers for inflation, visualize, and attempt to uncover insights from the data. TL;DR The top 3 revenue generating films (inflation-adjusted) are the first movie for each trilogy Disney era films do not make it past week 20 compared to the Lucas era On average, Lucas era movies generate 80% of their revenue within the first 10 weeks of release while Disney takes 2.8 weeks Load modules import pandas as pd import matplotlib.pyplot as plt import seaborn as sns sns.set(style="darkgrid") Define function def movie_revenue(era, trilogy, movie, url): # RETRIEVE DATA FROM URL movie_data = pd.read_html(url)[0] # CUMULATIVE REVENUE: TRANSFORM TO FLOAT AND CALCULATE PER MILLION (1e6) movie_data['Cumulative_Revenue'] = movie_data['To Date'].str.replace(',', '').str.replace('$', '').astype(int) movie_data['Cumulative_Revenue'] = movie_data['Cumulative_Revenue'] / 1e6 # WEEKLY REVENUE: TRANSFORM TO PER MILLION (1e6) FLOAT movie_data['Weekly_Revenue'] = movie_data['Weekly'].str.replace(',', '').str.replace('$', '').astype(int) movie_data['Weekly_Revenue'] = movie_data['Weekly_Revenue'] / 1e6 # SELECT WEEK INDEX & REVENUE DATA movie_data = movie_data[['Week', 'Weekly_Revenue', 'Cumulative_Revenue']] # ADD ADDITIONAL COLUMNS movie_data['era'] = era movie_data['trilogy'] = trilogy movie_data['movie'] = movie return(movie_data) Set parameters # LIST: PRODUCER ERA, TRILOGY, MOVIE NAME, URL TO CRAWL sw_list = [['Lucas','Prequel','The Phantom Menace','https://www.boxofficemojo.com/release/rl2742257153/weekly/'], ['Lucas','Prequel','Attack of the Clones','https://www.boxofficemojo.com/release/rl2809366017/weekly/'], ['Lucas','Prequel','Revenuge of the Sith','https://www.boxofficemojo.com/release/rl2943583745/weekly/'], ['Lucas','Original','A New Hope','https://www.boxofficemojo.com/release/rl2759034369/weekly/'], ['Lucas','Original','The Empire Strikes Back','https://www.boxofficemojo.com/release/rl2775811585/weekly/'], ['Lucas','Original','Return of the Jedi','https://www.boxofficemojo.com/release/rl2792588801/weekly/'], ['Disney','Sequel','The Force Awakens','https://www.boxofficemojo.com/release/rl2691925505/weekly/'], ['Disney','Sequel','The Last Jedi','https://www.boxofficemojo.com/release/rl2708702721/weekly/'], ['Disney','Sequel','The Rise of Skywalker','https://www.boxofficemojo.com/release/rl3305145857/weekly/'], ['Disney','SW Story','Rogue One','https://www.boxofficemojo.com/release/rl2557707777/weekly/'], ['Disney','SW Story','Solo','https://www.boxofficemojo.com/release/rl1954383361/weekly/']] Retrieve and parse data star_wars = [] for m in sw_list: data = movie_revenue(m[0], m[1], m[2], m[3]) star_wars.append(data) star_wars = pd.concat(star_wars) Spot check from random import randint star_wars.iloc[randint(0,len(star_wars))] ## Week 10 ## Weekly_Revenue 0.762516 ## Cumulative_Revenue 514.326 ## era Disney ## trilogy Sequel ## movie The Rise of Skywalker ## Name: 9, dtype: object Inflation-adjusted revenue # MOVIE TITLE m = [] for t in sw_list: title = t[2] m.append(title) # YEAR OF MOVIE y = [2000, 2002, 2005, 1977, 1980, 1983, 2016, 2018, 2019, 2017, 2018] # INFLATION RATE (https://www.bls.gov/data/inflation_calculator.htm) i = [0.497, 0.433, 0.32, 3.254, 2.129, 1.588, 0.074, 0.027, 0.008, 0.052, 0.027] # JOIN ALL THREE LISTS inflation = list(zip(m, y, i)) # CONVERT TO PANDAS DATAFRAME inflation = pd.DataFrame(inflation) # CREATE & APPLY COLUMN NAMES labels = ["movie", "year", "inflation_rate"] inflation.columns = labels # COMBINE DATAFRAMES star_wars_adjusted = star_wars.merge(inflation, how='left', on='movie') # INFLATION-ADJUSTED REVENUE: WEEKLY & CUMULATIVE star_wars_adjusted['Adjusted Weekly Revenue'] = star_wars_adjusted['Weekly_Revenue'] * (1 + star_wars_adjusted['inflation_rate']) star_wars_adjusted['Adjusted Cumulative Revenue'] = star_wars_adjusted['Cumulative_Revenue'] * (1 + star_wars_adjusted['inflation_rate']) Final spot check star_wars_adjusted.iloc[randint(0,len(star_wars_adjusted))] ## Week 46 ## Weekly_Revenue 0.792809 ## Cumulative_Revenue 213.955 ## era Lucas ## trilogy Original ## movie A New Hope ## year 1977 ## inflation_rate 3.254 ## Adjusted Weekly Revenue 3.37261 ## Adjusted Cumulative Revenue 910.166 ## Name: 101, dtype: object What is the total ticket sales for the entire franchise? star_wars_adjusted.agg({'Adjusted Weekly Revenue':'sum'})[0] ## 6125.216885688999 Over $6.6 billion in gross revenue (inflation-adjusted) was made from the US box office alone. ...

One of my dogs was recently diagnosed with an enlarged heart so the vet prescribed some medicine to mitigate the problem. The box came with a pamphlet which included the company’s effectiveness study for the drug, Vetmedin. I thought it would be fun to visualize one portion of the study with simulation. What follows is the #rstats code I used to examine and review the drug’s efficacy based on the reported results. ...

I moved to Los Angeles with my wife in October 2019 where we had a list of home improvement projects we wanted to complete or things to purchase. The problem we faced was disagreement on where to start since we had to juggle costs and compromise on what was most important at the time. For example, if we focused too much on lower ticket purchases we would delay projects that had potential to improve our home value. ...

Google has an amazing #rstats package called CausalImpact to predict the counterfactual: what would have happened if an intervention did not occur. This is a quick technical post to get someone up and running rather than a review of its literature, usage, or idiosyncrasies Load libraries library(tidyverse) library(CausalImpact) Download (dummy) data df <- read_csv("https://raw.githubusercontent.com/Eeysirhc/random_datasets/master/cimpact_sample_data.csv") df %>% sample_n(5) ## # A tibble: 5 x 3 ## date experiment_type revenue ## <date> <chr> <dbl> ## 1 2020-04-02 control 309. ## 2 2020-05-05 experiment 257. ## 3 2020-02-29 control 928. ## 4 2020-03-13 control 467. ## 5 2020-03-02 experiment 35.0 Shape data Before we can run our analysis, the CausalImpact package requires three columns: ...

The team at Bing were generous enough to release search query data with COVID-19 intent. The files are broken down by country and state level granularity so we can understand how the world is coping with the pandemic through search. What follows is an exploratory analysis on how US Bing users are searching for COVID-19 (a.k.a. coronavirus) information. tl;dr COVID-19 search queries generally fall into five distinct categories: 1. Awareness 2. Consideration 3. Management 4. Unease 5. Advocacy (?) ...