Since starting my transition to data analytics I've created six different portfolio projects. Ranging from showcasing my skills in SQL, Python, Tableau, R, and Excel. I've had plenty of people ask me about my process on creating projects. For almost all of my projects I have a dedicated article describing how I use specific tools to analyze, clean, and visualize data. But today I'll be going into an overview on how I complete projects.
This article is mostly focused for those who are aspiring data analyst and need more guidance on creating a project for their portfolio.
Below is my process:
Find a question - The first step is to try to find a question I want to answer. For my weight lifting project I wanted to improve my strength training program to get better in all of my lifts. For someone else it could be analyzing their sleep scheduled to see how much sleep they need and what times work best for them. Or you could answer a business question for a company or industry you're interested in.
Collect the data - The next step is to collect the data. You can either create your own set of data like I did for my weightlifting project. Or you can find a given data set using public sources like government websites or Kaggle or Maven Analytics Data Playground.
Explore the data - now you can explore your data using any tool you want. For this step I suggest using a tool you're comfortable with. For me it's Excel where I usually do my initial analysis because it's the tool I'm most comfortable with. This stage isn't about any analysis but exploring the data.
What questions can I answer - for this step I try to think of questions I can ask about the data that may lead me to my overarching goal. For instance for my weight lifting project I was asking things like what happens if the reps are higher? Does that change my strength? Or what's the average weight I lift? This part is all about asking questions that help you solve your initial question
Determine the tool/s - for each portfolio project I like to focus on a specific tool. For my weight lifting project I focused on SQL, for my Google Capstone Project it was R and Tableau. Because I'm trying to showcase specific skills this is how I choose my tool. For most of my projects I have an analysis portion and then a data visualization portion. Both of which use different tools.
Cleaning the data - this next part it's all about cleaning the data with the specific tool I decided in the last section. This could be using SQL, Excel or Python. Either way this is about cleaning the data.
Analyze the data - using the same tool from last time or even a different one. I analyze the data. This is about me answering the questions I asked in step 4. And really trying to see any trends, patterns that arise from this data. Looking for outliers or anything interesting that could help answer my question.
Visualize the data - This next step is one of my favorite steps which is visualizing the data. Typically for this I first create all the different types of charts I can using a specific tool like Tableau, you can do this step in the analysis and exploring phase as well. After that I contemplate the question I'm trying to answer and the best way to showcase the data in order to answer the question. I also think about the audience and how to best display the data so anyone can understand it without needing a data background. I personally like to visualize data so that anyone who understands basic charts like line graphs, bar charts, pie charts can get a gist of the data. I also sketch out a rough draft of the dashboard. This is when I typically use a data visualization tool like Tableau to create my visuals. But I also use a tool called Figma to create the backgrounds and make a more cohesive dashboard though this part is optional.
Provide insights - after cleaning, analyzing, visualizing the data. The last step is to provide the insights. While I do provide a summary of the most important parts. The essential part of data analysis is what insights you have gained and what this can do to answer your question. For this you can either create a presentation using something like PowerPoint, write a LinkedIn post, or what I do is write a detailed article on my blog.
This is my general process for creating comprehensive and detailed portfolio projects. Of course this may be different for everyone but this is what I found works best for me. Hopefully this helps you think about how your own process. This also may be subject to change because as I learn more about data analytics I will adapt and update my methods.
If you want more ideas on portfolio projects check out my article which has 10 portfolio project ideas.
Комментарии