Lecture 3 Homework

Homework 3: Pokemon Data Analysis

The following homework contains three questions, all of which require you to apply what you've learned about exploratory data analysis and visualizations. There is also a short discussion prompt for you to complete at the end. As always, please ask any of the mentors if you need help with this homework.

import pandas as pd import matplotlib.pyplot as plt import seaborn as sns import numpy as np

data = pd.read_csv("/work/Pokemon(1).csv") data.head()

#int64

Nameobject

Bulbasaur

Ivysaur

Venusaur

VenusaurMega Venusaur

Charmander

What is the most common "Type 1" for all the Pokemon? Use the code cell below to figure this out.

(Hint: Use the function value_counts() to answer this question)

data['Type 1'].value_counts()

Water is the most common.

Using data from every Psychic type (meaning Type 1 OR Type 2 is Psychic), create a scatterplot graph for the "Attack" and "Special Attack" columns. What kind of a relationship do you see? Is it a strong or weak relationship? Positive or negative?

# First filter the dataset so that it just includes Psychic type Pokemon. # Then make your scatterplot. q2 = data[(data['Type 1'] == 'Psychic') | (data['Type 2'] == 'Psychic')] sns.scatterplot(q2, x='Attack', y='Sp. Atk')

What kind of relationship do you see?

Strong positive relationship

Using histograms, compare the distributions of "HP" for Pokemon that are and aren't legendary. What do you see here? Talk about the skewness, outliers, mode(s), etc.

(Hint: Use the hue argument for sns.histplot to specify whether or not a Pokemon is legendary)

sns.histplot(data, x='HP', hue='Legendary')

Normal distribution of normal Pokemons.

Discussion Prompt: ChatGPT — A Teacher’s Worst Nightmare or the Future of Education?

Read the medium article linked here and respond to one of the questions listed under 'Further Questions to Consider:'. There should be a little text bubble icon at the bottom of your screen. Click on it and type your response there. This is just intended to get you guys thinking about what's happening in the data science world right now!

I would say researchers should use chatGPT a little bit to make their research more readable. I just read a poorly wording piece of article.