About This Project

At first, this project was created to see when the best time was to buy cards. Now, that is still the case and I fully intend on fleshing this project out in the future, but during my search for useable APIS, I happened upon a popularity chart and thus this iteration was born. With that said, this project aims to shed a faint light on the relationship between popularity and the Pokemon depicted in the card and whether or not that drives prices, increases volatility, and if so: what the impact on those variables actually are and whether or not a deeper, more thorough search is warranted.

Information and Troubleshooting

• If the price-history-pokemon-cards.ipynb errors on the first Run All, be patient and click it again.

• As far as the tcgplayer-scraper.py is concerned, that's a bit more fickle but only insofar as the data entry is concerned. For that, you should monitor your tcgplayer-data.csv. There may be some values missed. If that happens, delete the row, and then re-add it through retrying the scraper. This usually happens with older cards.

• The scraper, as it currently stands (3/27) does not check for duplicate entries into the database. Be careful!

• If there are no values in the '1Y' chart for the card of your choice, the scraper will not work.

Data Sources

CSVs:

1. tcgplayer-data.csv (gathered by my own webscraper)
2. pokemon-favorites-list.csv

Web Scraper:

1. tcgplayer-scraper.py

Installation

Before getting started, you'll need the following for the .ipynb to be read:

• Python
• pandas
• numPy

And these for the tcgplayer-scraper.py:

• Selenium
• Beautiful Soup
• webdriver_manager

All of these can be found at step 4.

Running the Notebook

1. As of this submission, Python 3.13.0 was used and Python 3 is required. Ensure that you have it installed.

2. Open your GitBash, navigate to where you want your folder to be, and clone the repo from my GitHub:

git clone https://github.com/jared-joering/pokemon_tcg_projects.git

3. To set up your virtual environment and activate it: change into the recently downloaded directory, set up, and activate your virtual environment:

cd pokemon_tcg_projects

python -m venv venv

source venv/Scripts/activate

4. To make sure you have all the required packages run:

pip install -r requirements.txt

Note: Both this project and the web scraper were created and tested in Visual Studio Code with Jupyter Notebook on two machines. One was a laptop and another a desktop, both Windows.

Running the Web Scraper

You'll need FireFox. That's it!

Using the Web Scraper

To run the web scraper, you'll just need to run the script, then find the full-URL of the card you wish to enter (i.e. 'https://www.tcgplayer.com/product/201352/pokemon-sm-cosmic-eclipse-pikachu-secret?Condition=Near+Mint&Language=English&page=1'), and paste it when prompted. The aforementioned card URL has been purposefully left out of the database. Give it a try!

Custom Data Dictionary

tcgplayer-data.csv

Column Name	Description	data type
'Card'	Pokemon depicted in the card	object
'Set'	Set containing the card	object
'Number'	Set number	object
'Rarity'	Rarity of the card	object
'Condition'	The condition of the card	object
'Volatility'	A metric of price fluctuations	object
'3/18 to 3/24'	Price during the week of March 3/18 to 3/24	float64
...	...	...
'3/10 to 3/16'	Price during the week of March 3/10 to 3/16	float64

Note: Columns from '3/18 to 3/24' on to '3/10 to 3/16' represent weekly price ranges for the full year.

pokemon-favorites-list.csv

Column Name	Description	data type
Pokemon	Name of the Pokemon	object
Number of votes	How many votes the Pokemon got	Int64
Rank	The respective ranking for each Pokemon	Int64

Data Analysis Capstone Satisfied Features List

1. Loading Data

    * Read two .csv's ('tcgplayer-data.csv' and 'pokemon-favorites-list.csv')
    * Scraped data with a web scraper (tcgplayer-scraper.py)
   

2. Clean and operate on the data while combining them

    * Cleaned (removed 'Condition', 'Volatility', and 'Votes') and merged 'tcgplayer-data.csv' and 'pokemon-favorites-list.csv' into 'final_df'.
    * Then calculated two new metrics 'Price-to-Populariy Ratio' and 'Popularity-Adjusted Volatility' (PPR and PAV respectively)
    * Transformed data as well (i.e. splitting, exploding, and mathmatically)
   

3. Visualize / Present your data.

    * Created six graphs (including two subplots) utilizing seaborn, numpy, and matplotlib.
   

4. Best practices: Enhance your project to a higher tier that will impress employers and help other programmers understand your project. Choose 2 from this table

    * Utilized a virtual environment, included instructions here.
    * Built a custom Data Dictionary, included here.
   

5. Interpretation of your data

    * Annotated my code with markdown, wrote clear code comments.  README is well-written.  Everything else is cleaned and functional.