This week in the Data School is dashboard week. This is a week of projects where we receive a challenge each day that we must wrangle the data and build a dashboard in the same day.

MealDB

Up first is a challenge using the MealDB API, an online database of recipes and foods from around the world. The approach I took to this was to try and download as much data from the different packets and clean and wrangle the data to make it suitable for visualisation.

Step One: Trusty old Alteryx

To begin I would use Alteyx to download and parse the JSON data from the API site. The site has a number of different APIs ranging from ingredients, category, area etc. I would require all this data to give my exploratory data analysis the best foundations.

I began by parsing the ingredients list and then using the ingredients to parse the meals associated with all the ingredients. This provided me with unique IDMeal field and from there I was able to parse all the individual meals. I had tried to create a similar approach using the recipes beginning with a-z. However, as is often the case this didn’t return all the recipes, so I changed tact and achieved the required result on take 2.

 

JSON Parsing

 

Once the data was parsed it was simply a case of transform the data with a couple of transposes and a crosstab to get the data and in particular the ingredients and measures into the correct the format. Finally, I supplemented the data with countries and population data to create a couple of silly metrics to add a bit of depth the analysis later.

 

 

alteryx workflow part 2

Tableau’in Around

Now the data was saved in a Hyper file I opened up Tableau and looked into finding patterns in the data. This was pretty tough as there was only 285 records and it was mainly recipes so not much in the way of insight. I managed to create a recipes per capita measure and looked into examining contributions by area/country and which ingredients and categories were most popular. Finally, I added the meals that associated with the afore-mentioned filters paired with recipe list and instructions. Below is final product filtered by the UK > Eggs > Breakfast and selecting “English Breakfast”.

 

meal db dashboard

The Data School
Author: The Data School