Hello everyone,

My name is Samuel.

Last year, I was looking to get a place in Lagos, and one of the locations I explored was Yaba.

The house-hunting experience was quite stressful.

From looking for good houses to meeting agents, ridiculous prices and so much more.

Reminiscing on my experience, I decided to play around with Power BI and create a simple but informative and up to date dashboard showing details about houses in Yaba, Lagos State.

Before starting; I sought to answer a few questions

1. What is the distribution of property types?

2. What is the average price for each property type?

3. Which location in Yaba has the highest number of property listings?

4. How does the price vary with the number of bedrooms, baths, and toilets?

5. What locations are the most expensive and how is the price distributed in these locations?

Data Collection

The data was obtained from propertypro . ng via power query

Usually, I'd use Python to scrape the data with libraries like beautifulsoup, but this time, I wanted to do everything in Power BI.

I searched for how to fetch data from multiple pages and found a good resource on Youtube.

This was the first new thing I learnt from this project. It was fun, and I might have use for it in the future.

Another good thing is the workbook maintains an active connection to the site, and data can be refreshed to get the current information.

Data Preprocessing

After bringing in the data, I found a lot of data cleaning to do.

Typically, I approach cleaning column by column from left to right.

reprocessing steps involved removing unnecessary whitespace in text columns, formatting text to follow the same pattern, identifying and removing sponsored adverts, fixing datatypes, outliers, and so much more.

The dataset looked something like this after cleaning in power query.



Moving on to the power bi frontend, I realised I needed to modify the types based on the number of rooms.

Using DAX switch statements, I categorised the rooms into duplexes, mini flats, self contain, two and three-bedroom flats

I also wanted unique locations. This was quite tricky, but I found the CONTAINSTRING function and combined it with a SWITCH statement to group properties by location.

The unique locations were listed on propertypro's website.

The final data looked like this.

Dashboard Inspiration & design

So I'm no designer, and I needed a dashboard inspiration.

I headed to Dribbble and searched for dashboards. I came across beautifully designed dashboards.

I picked one after much deliberation

Then I saved the picture.

To get the exact colours used in the project, I headed to color . adobe . com and uploaded the picture.

Finally, I headed to Figma and tried to create something similar to serve as a background for the dashboard's canvas.

I came up with this

well done