MethodCalls_EsterPark.py

import pandas as pd
import numpy as np
from collections import defaultdict
import matplotlib.pyplot as plt
from math import pi
get_ipython().magic('matplotlib inline')
fname = 'vgsales.csv'
x1 = pd.read_csv(fname)
"""
author: ester park
tried my best to use all style asserts like docstrings and parameter descriptions. 
used concepts covered in class for parsing the dataframes. 
mainly, functions, loops, classes, dictionaries, tuples, zips, lambda, 
list comprehension, conditionals/if-elif, numpy, various library usages 
(math, matplotlib), collections defaultdict
"""


class Strategies(object):
    
    def __init__(self, x1): 
        assert isinstance(x1, pd.core.frame.DataFrame)
        assert 'Year' in x1
        
        self.x1 = x1
        
    """
    author: ester park
    similar to our intervals hw7, i created a class that will accept
    a singular object that will be the dataframe dataset called self.x1
    return *this, will return self and output a scatter plot of 
    revenue in millions ($ USD million) vs. time (year)
    """
    def strategy1(self):
        """
        this will count() method will take in self.x1 which was the csv file given in week8 
        (not the new one)
        is has to parse the following series columns in order to project sales vs year
        Rank	Name	Platform	Year	Genre	Publisher	NA_Sales	EU_Sales	
        JP_Sales	Other_Sales	Global_Sales
        """
        
        
        yr = self.x1['Year']
        years = np.unique(self.x1['Year'][~np.isnan(self.x1['Year'])]).astype(int)
        NA_Sales = []
        EU_Sales = []
        JP_Sales = []
        Other_Sales = []

        for y in years: 
            NA_Sales.append(np.sum(self.x1[self.x1['Year'] == y]['NA_Sales']))
            EU_Sales.append(np.sum(self.x1[self.x1['Year'] == y]['EU_Sales']))
            JP_Sales.append(np.sum(self.x1[self.x1['Year'] == y]['JP_Sales']))
            Other_Sales.append(np.sum(self.x1[self.x1['Year'] == y]['Other_Sales']))

        plt.figure(figsize=(10,5))
        xaxis = range(len(years))
        plt.grid(True)
        plt.plot(NA_Sales, 'bo', label="NA sales")
        plt.plot(EU_Sales, 'ro', label="EU sales")
        plt.plot(JP_Sales, 'go', label="JP sales")
        plt.plot(Other_Sales, 'co', label="other sales")
        plt.ylabel("Sales in millions $")
        plt.legend(loc="upper left")
        plt.xticks(xaxis,years, rotation=70)
        a= plt.show()
        return a
    

    """
    author: ester park
    similar to our intervals hw7, i created a class that will accept
    a singular object that will be the dataframe dataset called self.x1
    return *this, will return plot of total revenue vs. number of games sold per company
    this explains the strategy for whether or not a company should invest their efforts
    in creating as many games as possible or specialize
    """
    def strategy2(self):
        """
        y axis = revenue in $usd millions
        x axis = number of games sold
        this function will take that same csv file and plot total revenue vs. number of 
        games sold per company the blue dots represent the other comapnies that are 
        part of the lower half of the outliers
        """
        publishers = self.x1.Publisher.unique()

        data = []

        max_pub = ''
        max_rev = 0 

        max_games_pub = ''
        max_games = 0
        for pub in publishers: 
            x = len(self.x1[self.x1['Publisher']==pub])
            y = np.sum(self.x1[self.x1['Publisher']==pub]['Global_Sales'])

            data.append((x,y, pub))

            if y > max_rev: 
                max_rev = y
                max_pub = pub

            if x > max_games: 
                max_games = x
                max_games_pub = pub

        plt.figure(figsize=(10,5))
        for x,y,pub in data: 
            if x>600: 
                plt.plot(x,y,'o', label=pub)
            else: 
                plt.plot(x,y,'bo')
        plt.legend(loc="upper left")
        plt.xlabel("Number of games sold")
        plt.ylabel("Total Revenue")
        plt.grid(True)
        return plt.show()
    

    def strategy3(self): 
        """
        y axis = Revenue
        x axis = number of games sold
        this function will take that same csv file and plot total revenue vs. number 
        of games sold per company however, we also want all the dots, 
        such that ea dot represents the number of games sold that year
        """
        publishers = self.x1.Publisher.unique()

        data = []


        for pub in publishers: 
            x = len(self.x1[self.x1['Publisher']==pub])
            y = np.sum(self.x1[self.x1['Publisher']==pub]['Global_Sales'])

            data.append((x,y, pub))


        data.sort(key=lambda tup: tup[1], reverse=True)

        top_10 = [d[2] for d in data[:10]]


        years = np.unique(self.x1['Year'][~np.isnan(self.x1['Year'])]).astype(int)

        publisher_data = {}
        for p in top_10: 
            publisher_data[p] = []
            for yr in years: 
                x = len(self.x1[(self.x1['Publisher']==p) & (self.x1['Year']==yr)])
                y = np.sum(self.x1[(self.x1['Publisher']==p) & ( self.x1['Year']==yr)]['Global_Sales'])
                publisher_data[p].append((x,y))



        plt.figure(figsize=(12,10))
        for p in publisher_data:
            x,y = map(list, zip(*publisher_data[p]))
            plt.plot(x,y, 'o', label=p)
        plt.legend(loc="upper left")
        plt.xlabel("Number of games sold")
        plt.ylabel("Total Revenue")
        plt.grid(True)
        return plt.show()
    

    """
    author: ester park
    will create a bar chart for each top 10 companies, what their avg revenue per year is
    thus signifying that even though nintendo made a lot more money than the rest,
    that on a yearly rate, they dont make the same rate as the other ones on avg
    which can be hurtful for quarterly investments/dividends
    """
    def strategy4(self): 
        publishers = self.x1.Publisher.unique()

        data = []


        for pub in publishers: 
            x = len(self.x1[self.x1['Publisher']==pub])
            y = np.sum(self.x1[self.x1['Publisher']==pub]['Global_Sales'])

            data.append((x,y, pub))


        data.sort(key=lambda tup: tup[1], reverse=True)

        top_10 = [d[2] for d in data[:10]]


        years = np.unique(self.x1['Year'][~np.isnan(self.x1['Year'])]).astype(int)

        publisher_data = {}
        for p in top_10: 
            publisher_data[p] = []
            for yr in years: 
                #x = len(self.x1[(self.x1['Publisher']==p) & (self.x1['Year']==yr)])
                yearly_revenue = np.sum(self.x1[(self.x1['Publisher']==p) & ( self.x1['Year']==yr)]['Global_Sales'])
                publisher_data[p].append(yearly_revenue)

        data = [publisher_data[p] for p in top_10]

        plt.figure(figsize=(8,10))
        plt.boxplot(data)
        plt.xticks(np.array(range(len(top_10)))+1,top_10, rotation=90)
        plt.xlabel("Top 10 Companies")
        plt.ylabel("Yearly Global Revenue in millions $")
        plt.title('Yearly Global Revenue for Top 10 Companies')
        return plt.show()


    """
    author: ester park
    now for the actual radar, i want to add the data frames into the areas
    https://python-graph-gallery.com/390-basic-radar-chart/
    https://www.kaggle.com/typewind/draw-a-radar-chart-with-python-in-a-simple-way
    just testing it out here
    """
    def strategy5(self):
        assert isinstance(self.x1, pd.core.frame.DataFrame)
        assert 'Year' in self.x1

        #need a dictionary key pair for the various game consoles and their number of games sold
        platforms = np.unique(self.x1['Platform'])
        genres = np.unique(self.x1['Genre'])
        platform_genres_rev = {}
        plt_counts = []
        for p in platforms: 
            count = len(self.x1[self.x1['Platform'] == p])
            #print('{} = {}'.format(p, count))
            plt_counts.append((p, count))
            platform_genres_rev[p] = []
            for g in genres: 
                rev = np.sum(self.x1[(self.x1['Platform'] == p) & (self.x1['Genre'] == g)]['Global_Sales'])
                platform_genres_rev[p].append(rev)
        plt_counts.sort(key= lambda tup:tup[1], reverse=True)
        print("Top platforms by number of games:")
        for i, (p, c) in enumerate(plt_counts): 
            print("{}. {}, count = {}".format(i+1, p, c))
        
        #print(platform_genres_rev['DS'])
        
        
        platform_revs = []
        for pl, revs in platform_genres_rev.items(): 
            platform_revs.append((pl,np.sum(revs)))
    
        platform_revs.sort(key = lambda tup:tup[1], reverse=True)
        print("\nTop platforms by revenue:")
        for i, (p,r) in enumerate(platform_revs): 
            print("{}. {}, revenue = ${:7.2f} million".format(i+1, p, r))
            
        # ------- PART 1: Create background
        # number of games sold radar chart
        # number of variable
        #categories=list(df)[1:]
        categories = np.unique(self.x1['Genre'])
        N = len(categories)

        # What will be the angle of each axis in the plot? (we divide the plot / number of variable)
        angles = [n / float(N) * 2 * pi for n in range(N)]
        angles += angles[:1]

        
        # Initialise the spider plot
        plt.figure(figsize=(10,10))
        ax = plt.subplot(111, polar=True)

        # If you want the first axis to be on top:
        ax.set_theta_offset(pi / 2)
        ax.set_theta_direction(-1)

        # Draw one axe per variable + add labels labels yet
        plt.xticks(angles[:-1], categories)

        # Draw ylabels
        ax.set_rlabel_position(0)
        y_ticks = np.array(list(range(10,250,40)))

        plt.yticks(y_ticks, y_ticks.astype('str'), color="grey", size=15)
        plt.ylim(0,250)


        # ------- PART 2: Add plots

        # Plot each individual = each line of the data
        # I don't do a loop, because plotting more than 3 groups makes the chart unreadable

        # Ind1
        #values=df.loc[0].drop('group').values.flatten().tolist()
        values = platform_genres_rev['DS'].copy()
        values += values[:1]
        ax.plot(angles, values, linewidth=1, linestyle='solid', label="DS")
        ax.fill(angles, values, 'b', alpha=0.1)

        # Ind2
        #values=df.loc[1].drop('group').values.flatten().tolist()
        values = platform_genres_rev['PS2'].copy()
        values += values[:1]
        ax.plot(angles, values, linewidth=1, linestyle='solid', label="PS2")
        ax.fill(angles, values, 'r', alpha=0.1)


        #values=df.loc[2].drop('group').values.flatten().tolist()
        values = platform_genres_rev['PS3'].copy()
        values += values[:1]
        ax.plot(angles, values, linewidth=1, linestyle='solid', label="PS3")
        ax.fill(angles, values, 'y', alpha=0.1)

        # Add legend
        plt.legend(loc='upper right', bbox_to_anchor=(0.1, 0.1))
        plt.title("Top 3 by # of games", y=1.08)
        plt.show()


        # ------- PART 1: Create background
        # num games sold 4-6 radar
        # number of variable
        #categories=list(df)[1:]
        categories = np.unique(self.x1['Genre'])
        N = len(categories)

        # What will be the angle of each axis in the plot? (we divide the plot / number of variable)
        angles = [n / float(N) * 2 * pi for n in range(N)]
        angles += angles[:1]

        # Initialise the spider plot
        plt.figure(figsize=(8,8))
        ax = plt.subplot(111, polar=True)

        # If you want the first axis to be on top:
        ax.set_theta_offset(pi / 2)
        ax.set_theta_direction(-1)

        # Draw one axe per variable + add labels labels yet
        plt.xticks(angles[:-1], categories)

        # Draw ylabels
        ax.set_rlabel_position(0)
        y_ticks = np.array(list(range(10,250,40)))

        plt.yticks(y_ticks, y_ticks.astype('str'), color="grey", size=15)
        plt.ylim(0,250)


        # ------- PART 2: Add plots

        # Plot each individual = each line of the data
        # I don't do a loop, because plotting more than 3 groups makes the chart unreadable

        # Ind1
        #values=df.loc[0].drop('group').values.flatten().tolist()
        values = platform_genres_rev['Wii'].copy()
        values += values[:1]
        ax.plot(angles, values, linewidth=1, linestyle='solid', label="Wii")
        ax.fill(angles, values, 'b', alpha=0.1)

        # Ind2
        #values=df.loc[1].drop('group').values.flatten().tolist()
        values = platform_genres_rev['X360'].copy()
        values += values[:1]
        ax.plot(angles, values, linewidth=1, linestyle='solid', label="X360")
        ax.fill(angles, values, 'r', alpha=0.1)


        #values=df.loc[2].drop('group').values.flatten().tolist()
        values = platform_genres_rev['PSP'].copy()
        values += values[:1]
        ax.plot(angles, values, linewidth=1, linestyle='solid', label="PSP")
        ax.fill(angles, values, 'y', alpha=0.1)

        # Add legend
        plt.legend(loc='upper right', bbox_to_anchor=(0.1, 0.1))
        plt.title("Top 4-6 by # of games", y=1.08)

        plt.show()


        # ------- PART 1: Create background
         #num games sold 7-9/genre
        # number of variable
        #categories=list(df)[1:]
        categories = np.unique(self.x1['Genre'])
        N = len(categories)

        # What will be the angle of each axis in the plot? (we divide the plot / number of variable)
        angles = [n / float(N) * 2 * pi for n in range(N)]
        angles += angles[:1]

        # Initialise the spider plot
        plt.figure(figsize=(8,8))
        ax = plt.subplot(111, polar=True)

        # If you want the first axis to be on top:
        ax.set_theta_offset(pi / 2)
        ax.set_theta_direction(-1)

        # Draw one axe per variable + add labels labels yet
        plt.xticks(angles[:-1], categories)

        # Draw ylabels
        ax.set_rlabel_position(0)
        y_ticks = np.array(list(range(10,250,40)))

        plt.yticks(y_ticks, y_ticks.astype('str'), color="grey", size=15)
        plt.ylim(0,250)


        # ------- PART 2: Add plots

        # Plot each individual = each line of the data
        # I don't do a loop, because plotting more than 3 groups makes the chart unreadable

        # Ind1
        #values=df.loc[0].drop('group').values.flatten().tolist()
        values = platform_genres_rev['PS'].copy()
        values += values[:1]
        ax.plot(angles, values, linewidth=1, linestyle='solid', label="PS")
        ax.fill(angles, values, 'b', alpha=0.1)

        # Ind2
        #values=df.loc[1].drop('group').values.flatten().tolist()
        values = platform_genres_rev['PC'].copy()
        values += values[:1]
        ax.plot(angles, values, linewidth=1, linestyle='solid', label="PC")
        ax.fill(angles, values, 'r', alpha=0.1)


        #values=df.loc[2].drop('group').values.flatten().tolist()
        values = platform_genres_rev['XB'].copy()
        values += values[:1]
        ax.plot(angles, values, linewidth=1, linestyle='solid', label="XB")
        ax.fill(angles, values, 'y', alpha=0.1)

        # Add legend
        plt.legend(loc='upper right', bbox_to_anchor=(0.1, 0.1))
        plt.title("Top 7-9 by # of games", y=1.08)

        plt.show()

        # number of variable
        #categories=list(df)[1:]
        categories = np.unique(self.x1['Genre'])
        N = len(categories)

        # What will be the angle of each axis in the plot? (we divide the plot / number of variable)
        angles = [n / float(N) * 2 * pi for n in range(N)]
        angles += angles[:1]

        # Initialise the spider plot
        plt.figure(figsize=(8,8))
        ax = plt.subplot(111, polar=True)

        # If you want the first axis to be on top:
        ax.set_theta_offset(pi / 2)
        ax.set_theta_direction(-1)

        # Draw one axe per variable + add labels labels yet
        plt.xticks(angles[:-1], categories)

        # Draw ylabels
        ax.set_rlabel_position(0)
        y_ticks = np.array(list(range(0,300,50)))

        plt.yticks(y_ticks, y_ticks.astype('str'), color="grey", size=15)
        plt.ylim(0,300)


        # ------- PART 2: Add plots

        # Plot each individual = each line of the data
        # I don't do a loop, because plotting more than 3 groups makes the chart unreadable

        # Ind1
        #values=df.loc[0].drop('group').values.flatten().tolist()
        values = platform_genres_rev['PS2'].copy()
        values += values[:1]
        ax.plot(angles, values, linewidth=1, linestyle='solid', label="PS2")
        ax.fill(angles, values, 'b', alpha=0.1)

        # Ind2
        #values=df.loc[1].drop('group').values.flatten().tolist()
        values = platform_genres_rev['X360'].copy()
        values += values[:1]
        ax.plot(angles, values, linewidth=1, linestyle='solid', label="X360")
        ax.fill(angles, values, 'r', alpha=0.1)


        #values=df.loc[2].drop('group').values.flatten().tolist()
        values = platform_genres_rev['PS3'].copy()
        values += values[:1]
        ax.plot(angles, values, linewidth=1, linestyle='solid', label="PS3")
        ax.fill(angles, values, 'y', alpha=0.1)

        # Add legend
        plt.legend(loc='upper right', bbox_to_anchor=(0.1, 0.1))
        plt.title("Top 1-3 by sales", y=1.08)

        plt.show()


        #amt of sales/genre/platform
        # number of variable
        #categories=list(df)[1:]
        categories = np.unique(self.x1['Genre'])
        N = len(categories)

        # What will be the angle of each axis in the plot? (we divide the plot / number of variable)
        angles = [n / float(N) * 2 * pi for n in range(N)]
        angles += angles[:1]

        # Initialise the spider plot
        plt.figure(figsize=(8,8))
        ax = plt.subplot(111, polar=True)

        # If you want the first axis to be on top:
        ax.set_theta_offset(pi / 2)
        ax.set_theta_direction(-1)

        # Draw one axe per variable + add labels labels yet
        plt.xticks(angles[:-1], categories)

        # Draw ylabels
        ax.set_rlabel_position(0)
        y_ticks = np.array(list(range(0,300,50)))

        plt.yticks(y_ticks, y_ticks.astype('str'), color="grey", size=15)
        plt.ylim(0,300)


        # ------- PART 2: Add plots

        # Plot each individual = each line of the data
        # I don't do a loop, because plotting more than 3 groups makes the chart unreadable

        # Ind1
        #values=df.loc[0].drop('group').values.flatten().tolist()
        values = platform_genres_rev['Wii'].copy()
        values += values[:1]
        ax.plot(angles, values, linewidth=1, linestyle='solid', label="Wii")
        ax.fill(angles, values, 'b', alpha=0.1)

        # Ind2
        #values=df.loc[1].drop('group').values.flatten().tolist()
        values = platform_genres_rev['DS'].copy()
        values += values[:1]
        ax.plot(angles, values, linewidth=1, linestyle='solid', label="DS")
        ax.fill(angles, values, 'r', alpha=0.1)


        #values=df.loc[2].drop('group').values.flatten().tolist()
        values = platform_genres_rev['PS'].copy()
        values += values[:1]
        ax.plot(angles, values, linewidth=1, linestyle='solid', label="PS")
        ax.fill(angles, values, 'y', alpha=0.1)

        # Add legend
        plt.legend(loc='upper right', bbox_to_anchor=(0.1, 0.1))
        plt.title("Top 4-6 by sales", y=1.08)

        plt.show()



         #amt of sales/genre/platform
        # number of variable
        #categories=list(df)[1:]
        categories = np.unique(self.x1['Genre'])
        N = len(categories)

        # What will be the angle of each axis in the plot? (we divide the plot / number of variable)
        angles = [n / float(N) * 2 * pi for n in range(N)]
        angles += angles[:1]

        # Initialise the spider plot
        plt.figure(figsize=(8,8))
        ax = plt.subplot(111, polar=True)

        # If you want the first axis to be on top:
        ax.set_theta_offset(pi / 2)
        ax.set_theta_direction(-1)

        # Draw one axe per variable + add labels labels yet
        plt.xticks(angles[:-1], categories)

        # Draw ylabels
        ax.set_rlabel_position(0)
        y_ticks = np.array(list(range(0,300,50)))

        plt.yticks(y_ticks, y_ticks.astype('str'), color="grey", size=15)
        plt.ylim(0,300)


        # ------- PART 2: Add plots

        # Plot each individual = each line of the data
        # I don't do a loop, because plotting more than 3 groups makes the chart unreadable

        # Ind1
        #values=df.loc[0].drop('group').values.flatten().tolist()
        values = platform_genres_rev['GBA'].copy()
        values += values[:1]
        ax.plot(angles, values, linewidth=1, linestyle='solid', label="GBA")
        ax.fill(angles, values, 'b', alpha=0.1)

        # Ind2
        #values=df.loc[1].drop('group').values.flatten().tolist()
        values = platform_genres_rev['PSP'].copy()
        values += values[:1]
        ax.plot(angles, values, linewidth=1, linestyle='solid', label="PSP")
        ax.fill(angles, values, 'r', alpha=0.1)


        #values=df.loc[2].drop('group').values.flatten().tolist()
        values = platform_genres_rev['PS4'].copy()
        values += values[:1]
        ax.plot(angles, values, linewidth=1, linestyle='solid', label="PS4")
        ax.fill(angles, values, 'y', alpha=0.1)

        # Add legend
        plt.legend(loc='upper right', bbox_to_anchor=(0.1, 0.1))
        plt.title("Top 7-9 by sales", y=1.08)

        return plt.show()