Analysis
8 Handy Python Libraries for Formatting and Cleaning Data
These Python libraries will make the crucial task of data cleaning a bit more bearable—from anonymizing datasets to wrangling dates and times.
Editor’s picks in Analysis
67 Years of Lego Sets
July 21, 2016• 8 minute read
There are many of us—errr, people—who build stuff with Legos at all ages. Having grown up with loads of hand-me-down Legos (and having a Lego Wall-E sitting on my desk right now), I started to wonder how Legos evolved from the sets I remember from my childhood to what they are today.
Joel Carron
Data Scientist at Mode
How DataHoliks Predicted Highfive's Customer Engagement
June 16, 2016• 7 minute read
In our previous blog post, we covered some of the exploratory analysis we did for Highfive, a modern meeting room collaboration and video conferencing solution. We determined the right metric for measuring customer engagement: something we call “sustained call hours/device/day” (SCH/BD). It’s a moving average of the hours of Highfive usage per day normalized by the number of Highfive devices.
Sourav Dey
Chief Data Scientist
Measuring Customer Engagement: How DataHoliks Helped Highfive Find the Right Metric
May 26, 2016• 7 minute read
Head into a conference room at RueLaLa or Expensify and you’ll probably see something special: people actually enjoying meetings. They’ve got Highfive, makers of a modern meeting room collaboration and video conferencing solution, to thank. The Highfive hardware installs right on top of a TV and pairs with cloud-based apps installed on employee computers, tablets or smartphones. Every employee on a domain is able to wirelessly screen share or join a video meeting from any device. Instead of per-user fees, companies choose a specific pricing bundle and pay per room based on their subscription choice.
Sourav Dey
Chief Data Scientist
