What is Health Informatics
26 March 2022
This has become a common question: "How do I create a data analyst portfolio?" I wish there were a one- or two-sentence answer to this, but unfortunately, there isn't. Thus, this article aims to highlight the few things that will make your analytic data portfolio stand out.
Since no one solution fits all to this problem, thus in this article, I'm going to try my best to break it down into four questions. This will help you gain some more clarity, and then you can come up with your action plans. Furthermore, throughout this article, you will find some tips, resources, and ideas to spark your creativity.
To answer this question, I would say that "it depends." This can vary widely depending on the type of analytics position you are seeking, the domain or industry vertical, the job description of the target organization, and so on. Perhaps you prefer data visualization over back-end data modeling and engineering. In any case, this article is designed for candidates interested in data visualization.
First, let's go to LinkedIn and pull up a few job postings regarding data analytics to see if there is any pattern there. It appears that the few data analyst and analytics jobs we examined share the following characteristics:
To demonstrate our skills, let's take a look at practical examples of each of these job descriptions. The first project idea is;
Data visualization is the visualization of information and data in a graphical form. Using visual elements like charts, graphs, and maps, data visualization tools provide an accessible way of viewing and understanding trends, outliers, and patterns in data. This form of data visualization helps to reach decisions faster and enables viewers to gain far better insights into patterns and trends.
Perhaps you are wondering about the difference between a dashboard and data visualization. A dashboard is a snapshot, or summary, of a large body of data. This is in contrast to data visualization, which presents data in a visual form. However, these two are often used in conjunction with one another.
Regardless of the industry, teamwork is a compelling concept. Building a team strategically for a highly significant project is an effective way to solve complicated challenges. Community service team-building experiences improve communication between employees, build stronger teams, positively impact the community, and motivate employees to strive for more incredible achievements. In terms of community participation in data science-related tasks, my top choice is Omdena.
Creating intelligent, actionable insights from data is the primary purpose of data storytelling. Data is brought to life through compelling narratives. Such narratives can lead to insights and "Aha" moments. An essential aspect of telling a business story through data is that the data can be tailored to the target audience, making it meaningful and appropriate for them.
Overall, we identified four types of projects that should be included in our portfolio. At this point, you may wonder how you can begin to work on such projects. You have a variety of options available to you. You can, for example, come up with a theory and then find a dataset to demonstrate your idea or vice versa. This ultimately leads to the next question.
It turns out that there are an enormous number of public datasets available. Kaggle is one of the most popular platforms for obtaining readily accessible datasets. Several other data sources are available, including Havard Dataverse, UCI Machine Learning Repository, and Google Public Data Explorer. Since these are all public datasets, they have already been cleaned, processed, and nicely formatted. Therefore, my recommendation would be to start by creating your dataset. In this way, your portfolio will gain a higher level of credibility. In addition, it will be able to demonstrate to recruiters that you can work with real-world datasets.
In order to create your data, you can either start with your personal data, such as the number of hours you spend reading every day, your daily or monthly expenditure, the number of calories you consume, and so forth. Another way to obtain datasets from the real world is to scrape data from the internet. You can find plenty of resources for data scraping, and I am adding a few links below to help you get started.
Another way of impressively getting data is by making use of APIs. There are a lot of companies that are generous enough to open up their APIs to the public. As APIs are now an integral part of modern data analytics daily activities, using APIs to grab the data for your portfolio will reinforce the credibility of your skills.
Here are some helpful links to help you find public datasets and APIs.
Since we now know how to acquire datasets for our projects, we will have to figure out which tools, libraries, platforms, and programming languages to utilize for data analysis and visualization purposes.
On a daily basis, data analysts are required to analyze, interpret, and visualize large datasets. Therefore, they must have the proper data visualization tools available at their disposal to carry out their work. The right tool makes it easier for data analysts to communicate their findings to a broader audience.
It is a tool used to visualize data and perform data analysis for business intelligence and data analytics. Tableau has been ranked as a leader in Gartner's Magic Quadrant for analytics and business intelligence.
QlikView is a classic solution for guided analytics. The software enables you to develop and deliver interactive analytics-based applications and dashboards in a short period of time. In a similar way to QlikView, Qliksense Desktop is an application that allows users to create visualizations, charts, interactive dashboards, and analytics apps that can be used locally and online. The key difference between QlikView and Qlik Sense is that QlikView is viewed as a first-generation analytics platform, whereas Qlik Sense offers modern analytical tools.
In short, Power BI is a collection of software services, apps, and connectors designed to help you transform unrelated data sources into meaningful, engaging, and visually immersive insights.
As you can see, the list goes on and on. Some of the products include data wrapper, plotly, sisense, etc.
Matplotlib is a Python plotting library that generates plots to allow them to be embedded into Python projects or program files using a Python object-oriented API.
Seaborn is a Java library that provides a high-level interface for drawing attractive and informative statistical graphics. If you read the official documentation on Seaborn, you will find that it is defined as a data visualization library built on Matplotlib. Simply put, Seaborn is an extension of Matplotlib that enables advanced plotting capabilities.
Bokeh is a Python library for building interactive statistical plots with simple commands. It allows for building complex statistical plots with simple commands in a matter of minutes, which makes it an excellent tool for interactive data visualization. Among the many reasons for Bokeh's popularity and its place in my list is its ability to visually enchant the user, causing them to focus on a particular aspect of the visualization. Additionally, Bokeh works very well with D3.js for creating interactive visualizations with a high interactivity quotient, even when dealing with enormous datasets or live streams.
One of R's rich ecosystem's most popular and commonly used visualization packages is ggplot2. It is a powerful package for creating beautiful graphs and plots within R.
In addition to Plotly (as we covered in Python), Plotly is also available in R. R's implementation of Plotly, allows the creation of interactive, web-based plots that make use of plotly.js as the base library. The advantage of using this package is that it can build contour plots, candlestick charts, maps, and 3D charts, which are impossible to create with most packages in R.
There is nearly no limit to the number of programming languages that can be used for creating data visualizations. However, Python and R are by far the most popular and widely used by data scientists.
Finally, it is your time to compile the resources mentioned above and create a visually pleasing visualization and intuitive insights using the information you have gathered. Once you have completed your project, be sure to share it with the public.
There is no doubt that a portfolio website is a preferred way to display your projects. I personally think that making a website for your portfolio is the first step toward getting the attention of a potential employer and landing a new job. Having said that, you need first to make sure that you know how to design a unique portfolio and stand out. There is no need to worry if you have no experience with web development or design. You can find a variety of free and paid portfolio templates and design services online. In my article about creating a portfolio website, you can find more information.
Finally, here are some great portfolio websites that may give you some inspiration.