Data Analysis

How I adjusted sales demand forecast by developing an index using casual inference

Image from

2020 was a shock wave for the world, especially for existing machine learning models in production and their ability( or lack of) to forecast correctly future business demands. Businesses either profited from this drastic change( significantly or otherwise) or suffered from it. This made the results of demand forecast models overestimate or underestimate demand than usual.

An excellent approach to mitigate these estimations is to create an index that captures this drastic change and include it in the model forecasts.

It is good to note that the 2020 shock wave affected businesses at different magnitudes and periods in that year…

How I use a bash script I created to automate it


If you are like me, you learn by creating pet projects. is an excellent source of pet projects data ( among other things).

I also use e as a workspace to build my pet projects. Again, this is because it has everything I can want in an environment like a file directory, easy access to GitHub, a terminal, ability to seamlessly delete and create a python environment. Plus, all the magic and experimentation happens outside my local system.

They also have chat support in every new project environment if you wanna rant about something not working on Deepnote.



How to use Pulp python library as an optimization method to decide if dynamic pricing makes sense


This is a two-part case study where we define the optimization problem in , and we use the Pulp python library as a tool to solve the business problem in part two.


From the , we were able to define the decision variables, objective function, and constraints. Let have a quick refresher mathematically.

  • Decision variable
Our decision variables can be represented as A and Swhere 
A is the number of Apple phone stock
S is the number of Samsung phone stock
  • Objective function
Z = 788.53A + 425.81Swhere 
Z is the optimal maximum revenue
788.53 and 425.81…


Deciding how many smartphones to have for each brand and if dynamic pricing makes sense


This is a two-part case study where we define the optimization problem in part one, and we use the Pulp python library as a tool to solve the business problem in .


Optimization problems are the dilemma of any business. This can come in two decision-making: either maximation and minimization, be it profit or cost.

Any optimization problem has three things you need to define before solving them

  1. The decision variable s— what you can control
  2. The objective function — What we want to achieve
  3. Constraint(s) — What limited resources we have

A typical optimization problem can be solved…

How to use DataBricks widgets and other things in between

Image from


I recently started using DataBricks more frequently in the past 2 months, and I have had many mood swings about it. One minute I am super impressed, and the other minute, I abhor it. To be fair, a lot of the negative emotions about it can be linked to the fact that it needs some getting used to, and I find myself lacking the time to do that.

But what is DataBricks?

In simple terms, DataBrick is like Jupyter notebook but so much more and on steroids. It trumps Jupyter notebook in terms of

  • Being able to run spark SQL…

Articulating how others convey information through visualizations


Visualization is one of the cores of data analytics. It is one of my favourite part of analytics because it challenges your communication skills and is the most user-centric part of the analysis.

Even though my default programming language is python. It is not my favourite tools for visualization: Tableau takes that position. And for over a year, I have been consciously taking to improve how I convey information through visualization using Tableau most of the time.

If you check my , you will immediately realize that I am nowhere near perfect, and I still make a s**t load…

A book on career advice by Cal Newport

Image made by Author using Canva

I know, I know. The book has a cheesy title. At least that was my first impression. I assumed it was a marketing gimmick book with a “clickbait” title that would have very fluffy assumptions as content. If not because a trusted source recommended this book, I probably would not have been “caught dead” reading it. As they say, do not judge a book by its cover.

For one thing, the entirety of this book confirms the two things I am beginning to understand

  • Common sense is not very common.
  • Conventional ideologies can be very flawed.

For me, this book…

How to view your Tableau dashboard on Streamlit

Image by Markus Spiske on

As 2020 draws to its ends (hopefully along with its chaos), I took a step back to reminisce about all the tools/skills I have picked up this year and all the tools I really didn’t use so much this year. One tool I really didn’t use so much is Tableau.

Tableau is a powerful data visualization tool that creates visually-appealing dashboards.

I decided to find a dataset to analyze and visualize on Tableau.

One tool that I picked up this year and I have used quite frequently is Streamlit.

is such an awesome tool for data scientists.

Instead of…

How I converted my medium post to realistic Audio using AWS Polly

Image by James McKinven from


I recently discovered that I retain more information when I read a book in audio format than in written format. This discovery instigated the need to find ways to read Medium posts in audio formats like a podcast. My first plan was to gather all my medium read lists and convert them to audio. Unfortunately, Medium API technically does not support this. What it does support, however, is getting article feeds by username. Therefore I decided to streamline my goal, for now, to convert one of my old medium posts to audio. Here is the sample below

You can…

Deploy from your command line, manage your AWS instances, and more

Photo by on .

Amazon Web Services (AWS) is one of the market leaders in cloud computing. It has lots of services, ranging from infrastructure as a service (IAAS) to platform as a service (PAAS) and some software as a service (SAAS).

AWS services can be accessed in two ways: via the GUI (the AWS management console) and programmatic access via API, CLI, SDK, etc.

For this article, we will be focusing on setting up the CLI interface.

There are two major benefits of using the AWS CLI interface:

  • It saves time.
  • It aids the automation of workflows/processes within AWS services.

Setting Up


  • Create or have…

Okoh Anita

Data Scientist | Building Data Product with Python, Analytics, and Machine Learning | New Technologies Obsessed| Self-care Conscious

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store