Cooking can be fun but challenging. Finding the right recipe is not always easy.
Epicurious.com is a popular website for all types of recipes. I want to see if the tags are reliable and if I can find any indicators by looking at their recipes.

One of the features is the tag function. By tagging recipes, recipe publishers can give a bit more insight into the culture or background of the recipe. I will be analyzing their usage and accuracy.

Recipe data is collected by using BeautifulSoup and MongoDB. It is then converted to Pandas Dataframe for EDA. Simple iteration functions were built for this. I then looked through the columns and extracted what I needed. Cleaned nan data where necessary.

For each hypothesis, I would switch between filtering the recipe tags or ingredients. This involved simple pandas commands. To find my distributions, I would plot the density functions of the parameters I was checking. I noticed there were many outliers in the data so I had to clean the data.

After plotting the distribution we can see that it is not a normal distribution. Thus I used the U-test for my hypothesis tests.

Hypothesis Test 1: Trying to lose weight is common in the summer. I chose to look at data with summer and winter tags and compare the fat content. I performed a U-test on my hypothesis

H0: summer fat content is higher than winter. H1: winter fat is higher.

P<.02, Reject the null hypothesis (0.000398)

Although some of the data needed to be wrangled, I believe my hypothesis tests shows that the tags can be a reliable source when determining what kind of meals you want to cook. Whether it is a fatty winter dessert or a healthy summer/vegetarian dish, Epicurious is a great resource for delicious meals.