Module 4 - Tutorial - Data Visualization

import pandas as pd

from matplotlib import pyplot as plt

import seaborn as sns

%matplotlib inline

location = Pokemon.csv

df = pd.read_csv(location, encoding = "ISO-8859-1") #1st column needs encoding to read pound/hashtag(#) symbol

df.head()

#rename the pound/hashtag(#) column so it doesn't cause further issues

df.rename(columns={'#': 'Number'}, inplace =True)

#validate

df.head()

#set indices as Pokemon numbers

df.set_index('Number', inplace=True)

df.head()

#make a scatterplot

sns.lmplot(x='Attack', y='Defense', data=df)

#add more details to scatterplot

sns.lmplot(x='Attack', y='Defense', data=df, 

           fit_reg=False, #remove regression line

           hue='Stage')   #color by evolution stage

#boxplot showing stats for each characteristic

sns.boxplot(data=df)

#create a dataframe containing the stats for each Pokemon

#drop Total, Stage, and Legendary

stats_df = df.drop(['Total', 'Stage', 'Legendary'], axis=1)

stats_df.head()

#boxplot for stats

sns.boxplot(data=stats_df)

#change size of plot

plt.subplots(figsize=(13,7))

#violin plot shows the distribution of attack power for each type

sns.violinplot(x='Type 1', y='Attack', data=df)

#create color palette

#color Hex numbers are from Bulbapedia (https://bulbapedia.bulbagarden.net/wiki/Category:Type_color_templates)

pkmn_type_colors = ['#78C850',  # Grass

                    '#F08030',  # Fire

                    '#6890F0',  # Water

                    '#A8B820',  # Bug

                    '#A8A878',  # Normal

                    '#A040A0',  # Poison

                    '#F8D030',  # Electric

                    '#E0C068',  # Ground

                    '#EE99AC',  # Fairy

                    '#C03028',  # Fighting

                    '#F85888',  # Psychic

                    '#B8A038',  # Rock

                    '#705898',  # Ghost

                    '#98D8D8',  # Ice

                    '#7038F8',  # Dragon

                   ]

#violin plot with Pokemon color palette

plt.subplots(figsize=(13,7))

sns.violinplot(x='Type 1', y='Attack', data=df, palette=pkmn_type_colors)

#swarm plot

plt.subplots(figsize=(10,6))

sns.swarmplot(x='Type 1', y='Attack', data=df, palette=pkmn_type_colors)

#combine a violin plot with a swarm plot

plt.figure(figsize=(13,7))

sns.violinplot(x='Type 1',

               y='Attack',

               data=df,

               inner=None, #removes bars inside violin plot

               palette=pkmn_type_colors)

sns.swarmplot(x='Type 1',

              y='Attack',

              data=df,

              color='k', #make points black

              alpha=0.7) #slightly transparent

plt.title('Attack by Type')

#create a column that contains all stat types and a column for their corresponding value

melt_df = pd.melt(stats_df, 

                  id_vars=['Name', 'Type 1', 'Type 2'], #column to keep

                  var_name="Stat")

melt_df.head()

#melted dataframe has 6 times the amount of rows as original stats dataframe

#6 rows for each pokemon for each stat type

print(stats_df.shape)

print(melt_df.shape)

plt.figure(figsize=(10,6))

#swarm plot of Stats from melted dataframe

sns.swarmplot(x='Stat', y='value', 

              data=melt_df,

              hue='Type 1')

plt.figure(figsize=(13,7))

sns.swarmplot(x='Stat', y='value',

              data=melt_df,

              hue='Type 1', 

              dodge=True, #separate points on chart by hue

              palette=pkmn_type_colors)

#adjust the y-axis

plt.ylim(0,260)

#put the legend on the right

plt.legend(bbox_to_anchor=(1,1), loc=2)

#show correlation of stats via heatmap

corr = stats_df.corr()

sns.heatmap(corr, vmin=-1, annot=True)

#create a histogram of the distribution of attack power

sns.distplot(df['Attack'])

#create a bar plot of Type 1 frequency

sns.countplot(x='Type 1',

              data=df, 

              palette=pkmn_type_colors)

plt.xticks(rotation=-45)

#scatterplot by each stage type (evolution level)

g = sns.factorplot(x='Type 1', y='Attack',

                   data=df,

                   hue='Stage', 

                   col='Stage',

                   kind='swarm')

#rotate x-axis labels on ALL charts

g.set_xticklabels(rotation=-45)