As a training for myself to get more familiar with image classification and OpenCV I followed the tutorial on http://www.pyimagesearch.com/2014/08/04/opencv-python-color-detection/. I modified it slightly to fit my environment.

In my previous post on testing in Python I explained how to test in a Jupyter Notebook. Now a few weeks later I need to get familiar with mocking in Python. Since this is a new concept for me I made a small notebook testing different functionalities of the mocking library.

Unittesting in a Jupyter notebook

To add some custom style to your dashboard, you can create a cascade stylesheet in your app folder and overrule the standard dashboard lay-out.

In this short tutorial I will create a Splunk app to monitor the webserver. We will create the bare-bone app and we add a few panels to the dashboard.

Since I have been using Linux for a while, switching back to the Windows CLI has some challenges. Luckily you can add aliases to the command prompt with the following procedure.

Installing the diff/merge tool KDiff3 is easy using the package manager Homebrew extension Cask. The extension makes is possible to install (GUI) applications on the Mac without the dragging and dropping of the DMG-files.

jitsejan@MBP $ /usr/bin/ruby -e "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install)"
jitsejan@MBP $ brew tap caskroom/cask
jitsejan@MBP $ brew cask install kdiff3

Explanation of the installation of Splunk.

To add some style to the notebook, Jupyter has the option to add a custom css file. This file should be located in ~/.jupyter/custom/custom.css. If this file does not exist yet, create the directory if needed and add the stylesheet. Add the following code and refresh the notebook to see if it works.

body {
  background-color: purple;
}

To make it easier to modify this file, create a link from the stylesheet to the working directory for the notebooks in order to be able to modify the stylesheet in the browser.

jitsejan@jjsvps:~/code/notebooks$ ln -s ~/.jupyter/custom/custom.css .

In this post I will try to make a world map to indicate where the 50 artists I listen to most on Spotify are from. This will be my first personal data science project, so bare with me. My expectation is that the majority will be in either the UK (indie bands), the US or from South-Korea (K-pop). Lets see if my hypothesis is true.