Previously, I have talked about some of the plans for improving the dashboard.
Gathering More Data
Currently, the python dashboard only focuses on the oil prices, oil production, and oil consumption for different countries. I would like to expand more on the different kind of data to display on the dashboard. The extra data will be gathered from US Bureau of Labor Statistics, they offer data finder within their website to help search for the data I want.
As we can see from the figure above, the website has data for prices regarding different types of food and commodities. This will be very useful since my current CPI dashboard only shows the CPI values, and it does not show how CPI values can impact food prices.
Let’s take a look at the eggs data from the figure above and see what the data finder can offer.
At the top, we have the option to select the date range, and changing the start year can help us select the date range that we want. On the bottom right of the picture above, there is a panel there showing some of the summaries for the data. The time range for this particular set of data goes as far back as 1980 to 2022.
This dataset is measuring the price of grade A, large dozen eggs by city average, so which means that some cities will have more or less expensive eggs per dozen. This data is also not seasonally adjusted for the effect of holidays, or any trends.
The following datasets are the ones that I would like to display on the dashboard:
- Electricity per KWH
- Eggs
- Milk
- Bacon
- Bread
- Flour
- Rice
- Chicken Breast
- Chicken (Whole)
- Bananas
- Ground Chuck
- Ground Beef
- Coffee
- Cookies
- Potato Chips
- Tomatoes
- Oranges
Next, I export the data from the website into CSV, and the CSV have the following structure:
Series ID: Corresponds to the different item in the dataset
Year: Year it was recorded
Period: What month of that particular year it was recorded
Label: Seems to be Year + Month
Value: Value of the a particular dataset during particular Year and Month
When I import the data for processing at the next step, I will most likely remove the Year and Period due to redundancy. Most likely, I would just separate Label into Year and Month. There are a total of 8006 records in the CSV for all the categories I mentioned above.
