Affiliate programs are amazing. All you have to do is point people to someone else’s amazing product and bam, you get a share of the cut. Many sites have affiliate programs, more than you would probably think. One of the behemoths in this space? Amazon. Amazon has such a wide variety of items to recommend as well as a solid and trustworthy reputation that it’s rather easy to get people to buy off Amazon.

However, Amazon has become increasingly more difficult to become accepted into their affiliate program. They now have a 14,867 word document detailing the Affiliate program policies. I built a site that garnered about 30,000 views, $400 in revenue, and then had it disappear overnight when Amazon rejected the application. Had this happen another time, until finally I was accepted into the program. …

Extracting meaning from unstructured data is a difficult thing to do. Sometimes, if you’re lucky, there are telling characteristics about the data that provide an interesting angle into the text. One angle is web addresses. Web addresses are semi-structured data that can be extracted to give some proxy of meaning to a body of text.

I took this approach to find trends and meaning in Reddit comments. The corpus of all Reddit comments from January 2015 to June 2017 live on Google BigQuery datasets. I extracted all links that contained amazon.com in them and then ensured they were a product by scraping the Amazon page. What I found was fun and interesting. Based on this premise, I built a site called ThingsOnReddit which organizes products by subreddit. The diversity of people on Reddit leads to all sorts of products being mentioned. So, what were the interesting things found on Reddit? …

Doing data analysis can be fun and rewarding. It’s something I do a lot of in my free time. Without the right tools though, it can be frustrating and extremely time consuming. I break down the process of working with data into 4 steps.

1. Data gathering: Finding and getting the dataset you are interested in
2. Data cleaning: Getting data into the proper format
3. Data exploration: Finding trends and interesting patterns
4. Data visualization: Visualizing the awesome trends you’ve found

Data gathering

The process of gathering data has gotten significantly better in the last 5 to 10 years. …

I recently moved from Boston to San Francisco. In the months leading up to the move, I scoured mailing lists and craigslist to find cheap apartments. The prices listed were astronomically high. It was common for me to see apartments listed for around $1000 to share a room in an apartment with 5 to 7 people. Due to rent control, it is still possible to find affordable rooms, but they are few and far between.

Just how high is rent in San Francisco? America? If you hate paying high rent prices for crappy rooms like me, below might be an indicator for which cities to live in, or which cities not to live in. …

Integrating Devise’s Google Omniauth into Rails and then utilizing that to access Google APIs seems like something that should be straightforward. Unfortunately, this task is surprisingly difficult as not all the pieces fit together perfectly, and the documentation to do it, is spread across the Internet in forums, Github Issues, and docs. The goal of this post is to allow you to build a minimal Rails 5.1 application that can authenticate with Google and then call Google APIs.

Assumptions

• Working knowledge of Rails and encrypted secrets
• Using Rails 5.1
• Familiarity with Devise

Device Authentication with Omniauth

Devise put together a nice tutorial on how to setup omniauth with facebook. I would recommend reading that to get an overview. Setting up omniauth with Google is very similar. …

This guide describes how to change the color of an icon for Photoshop CC 2017.

1. Open and select layer in Photoshop
2. Click the fx button in the lower right in the Layers panel

3. Choose Color Overlay

4. Choose Color and click OK