Data Visualization Project

Project Desctiption

This project uses a Pokemon Go dataset found on Kaggle (https://www.kaggle.com/datasets/shreyasur965/pokemon-go?resource=download) and creates a few different graphs/tables based on the information present in the dataset. When first loading the Visualization, it loads to Bulbasaur (Pokemon ID 1) and presents four different important bits of information. The first at the very top of the visualization, it shows the selected Pokemon's name, ID and typing. Moving to the top left, a radar chart shows how the pokemon's base stats compare to the max of all pokemon for each stat, as well as the actual numbers behind it. The top right has a graph that looks similar for most pokemon, but shows how the pokemon's CP grows as it levels up. On the default view, it will label the CP at levels 1, 10, 20, 30, and 51, however this can be changed to an absolute value where the top of the graph is determined by the pokemon with the highest CP. When the graph is in absolute mode, we only display the pokemon's max CP due to potential overlapping text. Lastly, there is a table on the bottom left that has Type effectiveness matchups as well as the moves it can learn. In the type effectiveness table, it also lists the precise effectiveness of moves against that pokemon. Note that the type effectiveness in Pokemon Go is very different from the mainline Pokemon games. Note how the lowest possible effective value (0.244x) has an asterisk to indicate that it is very ineffective, and bold and normal text are used to differentiate the type effectiveness otherwise. For more details on that, read more in my Week 4 Progress section. At the very top, there are two dropdown menus and a input box. The first dropdown menu selects a pokemon based on it's name. The second dropdown controls the graph on the top right's setting. The input box allows you to quickly type in a pokemon's name or ID number and it'll load that pokemon. Overall, I am happy with how this project turned out, but I still have a few changes I would have liked to make if I had more time.

Future Work

I was unable to find the time to make the dropdowns and input box at the top as pretty as I would've like, but as of now they are functional and that matters more than being pretty for it's purpose. In addition, Pokemon who have a Max CP equal to or close to the Max CP of the strongest pokemon (Currently Zacian, ID = 888) will still have an issue of overlapping text when in the absolute view on the level scaling graph. I'm sure there is a fix for this, but I cannot think of it at the moment. I also cut the "methods of obtaining pokemon" that was present in my original sketch because I'm not sure I had the ability to make those complicated shapes look good and still hold what they were supposed to be, and because those fields were often misleading or incorrect entirely for any given pokemon. Lastly, the moves do not contain any specific information on them. I would've loved to include more information on the moves, however I could not find a dataset that contains information on all the moves in Pokemon Go, so for now it will remain as a future work opportunity.

Weekly Progress

Week 1 Progress

Unfortunately, due to a slew of other projects and problems, I didn't really have time to work much on this step. I will continue to work as I am freed up.

Week 2 Progress

This week I managed to complete Week 1's tasks. Now the Selected Pokemon (default is Bulbasaur) will display the appropriate radar chart. Also, I removed 3 pokemon from the pokemon.csv since they had invalid data for their stats in these categories. These pokemon were ID = 679, 680, & 681. Currently, the page does not refresh properly when changing values in the viz, causing it to redraw over the old viz. This means that the whole page must be refreshed in order to see proper changes. I'm not currently sure what causes this, but I will investigate soon.

Week 3 Progress

This week, I decided to skip ahead to Week 4's tasks. Now, there is a dropdown menu where you can select a pokemon based on it's name. This dropdown menu populates with all pokemon names in the pokemon.csv file. In addition, I removed one extreme outlier (Eternatus, ID = 890) from the pokemon.csv, due to it's extremely high Max CP. I cleaned up the existing code to be slightly easier to work with in my opinion. Lastly, I solved the redraw bug that I had last week, as well as a bug related to the radar chart not drawing properly that I did not know about, which is part of the reason I decided to implement the dropdown menu this week. For next week, I hope to implement the Weaknesses/resistances & move pools, which should add some color to the visualization (I know that was this weeks assignment, but I felt this was more important to complete this week).

Week 4 Progress

This week, I implemented the movepool lists, matchups table, weaknesses/resistances, a level curve which can be scaled to be relative or absolute, and added some colors to the visualization. The movepool lists were relatively trivial, however I am unable to get information on individual moves, so I cannot change their color based on the type of the move listed (which is something I would have done if possible). For the Matchup table, I made a matrix 18x18 which contains information on every type's defenses versus an enemy attack of any type. This was mindless, but time consuming, as I had to input all 324 possibilities by hand. I also added a table which tracks a pokemon type with a corresponding hex color, which are the official type colorings used in the games & materials. Once I had finished that, I added colors to the types in the resistances and weaknesses lists, which also differentiates if a move is "super effective", "very effective" (Both types are weak to a given attacking type), "not very effective", "ineffective" (Either both types resist or one type is immune to a given attacking type), or "very innefective" (One type is immune, the other is resistantto a given attacking type) by using bold for Very effective, ineffective, and very ineffective, as well as adding an asterisk to the text if a pokemon has any "Very innefective" resistances.

As for the level curve, I was able to reuse the invlerp function I made for the radar chart which helped a lot. The formula is a lot to look at, but it wasn't too hard to implement since it is well documented online. I managed to complete the level curve, at which point I realized it's not as much of a curve as I would have guessed, but rather a compound function consisting of two lines (it's techically more, but there's a stark contrast between levels 1-30 and 30.5-51). I marked a line at level 30 to better show this, as well as other notable points on the graph. I also decided to use the secondary typing for the level curve if the pokemon has two types.

Lastly, As I was looking through the data more closely, I realized that for many pokemon which have alternate forms, it uses that alternate form instead of the original. For example, the pokemon Voltorb, which is normally just an electric type, is listed as an electric/grass type in the CSV. This also may effect the default stats of some pokemon, which can be seen in Mewtwo, which has the alternate form called "Armored Mewtwo". Regular Mewtwo has base 300 Attack, 182 Defense, and 214 Stamina, as compared to Armored Mewtwo which has base 182 Attack, 278 Defense, and 214 Stamina, which is what displays in the visualization. Unforunately, there are no ways to see the regular forms of pokemon in the csv I used, so I will have to settle for this in my visualization.

Week 5 Progress

Since I was pretty happy with my progress from last week, I used this week to make minor changes to the visualization. Firstly, I added the multiplier number itself on the weaknesses & resistances, for better clarity. In addition, in the relative scaling view of the level curve, I added more info on what the CP is of the pokemon at levels 10, 20, and 30. Lastly, I added a couple of functions that would help with redundant pieces of code in order to clean it up a small amount. I also fixed a minor errror with Bug type pokemon's weaknesses and resistances. For next week, I plan on adding a number entry box as an alternative to using the dropdown menu to find a pokemon.

Week 6 Progress

This week, I set out what I wanted to do by adding an input box at the top that will allow users to type in an ID or pokemon name. The pokemon name must be an exact match, however it does not check for capitalization. Names of pokemon including unusual characters, such as "Ho-Oh" and "Brute Bonnet", still worked as intended. As a reminder, some pokemon were removed due to missing stats or extreme outlier case. These Pokemon are ID numbers 679, 680, 681, and 890.

Also, as I tested some pokemon, I noticed that their moves (Fast/Charged attacks) were essentially dummy data, however this is again an issue I cannot fix easily with my data, especially since these pokemon haven't yet been made available in the official game.

Initial Proposal

Data

The data I propose to visualize for my project is a Pokemon GO dataset that can be found at https://www.kaggle.com/datasets/shreyasur965/pokemon-go?resource=download

Questions & Tasks

The following tasks and questions will drive the visualization and interaction decisions for this project:

• I want to see how many pokemon have special methods of finding them (i.e. found_egg, found_research etc.)
• I want to see how different pokemon's max strength (max_cp) correlates to other fields (base_capture_rate, etc.)
• I want to see how different pokemon's move sets vary across different stages of evolution.

Sketches

In this rough diagram of the dataset, I show different stats of a particular pokemon as a radar chart, a level curve which will display a pokemon's max strength at every level, and a few symbols which will display which pokemon can be captured by different methods. The different methods of obtaining come from eggs, evolution, wild captures, research breakthroughs, and raid battles. In this diagram I also leave space for different moves the pokemon can learn, as well as it's weaknesses and resistances.

Open Questions

My biggest worry is actually the dropdown/search for a particular pokemon. I intend for it to populate with pokmemon's names from the dataset, but I'm not confident about putting it all into a dropdown menu. In addition, the symbols I plan on using for the ways to obtain might not be as clear as I hoped, so I plan on iterating them over time to make sure that the symbol's meaning is clear. Lastly, the level curve is going to originate from a formula based on the stats of the pokemon, so I have to make sure that the formula is accurate to the game information.

Milestones

Week 1: Implement radar chart, ways to obtain, fast attack list, and charged attack list.

Week 2: Implement weaknesses & resistances lists

Week 3: Implement level curve

Week 4: Implement Dropdown to search different pokemon

Week 5: Catchup on any incomplete tasks, implement any last minute ideas.