Análisis de los 25 retailers más grandes de Estados Unidos

# Importar librerías aquí import pandas as pd import seaborn as sns import matplotlib.pyplot as plt import numpy as np from colors import *

# Importar datos aquí df = pd.read_csv('largest_us_retailers.csv') df

#Ordeno la lista para detectar el que más vendió df_sorted = df.sort_values('Sales',ascending=0) df_mean_sales = df_sorted[1:]['Sales'].mean() print(FG_BLUE) print('Promedio de ventas sin Walmart: %i USD'%int(df_mean_sales) )

minval = df_sorted.quantile(0) Q1 = df_sorted.quantile(0.25) median = Q1 = df_sorted.quantile(0.50) Q3 = df_sorted.quantile(0.75) maxval = df_sorted.quantile(1) print(minval,Q1,median,Q3, maxval)

from colors import * walmart = df_sorted.iloc[0]['Sales'] #ventas de walmart all_no_walmart = df_sorted[1:]['Sales'].sum() #ventas de todas menos walmart all_companies = all_no_walmart + walmart #ventas totales print(FG_BLUE) print('Ventas de Walmart: %i USD'%int(walmart) ) print(FG_RED) print('Ventas de todas las compañias (menos Walmart): %i USD'%int(all_no_walmart)) print(FG_GREEN) print('Ventas de todas las compañia: %i USD'%int(all_companies))

companies = df_sorted['Company'][1:][::-1] sales = df_sorted['Sales'][1:][::-1] fig = plt.figure(figsize = (10,10)) ax = fig.add_subplot() plt.stem(sales, orientation='horizontal',linefmt='k',markerfmt='C1o') ax.set_yticks([i for i in range(len(companies))]) ax.set_yticklabels(companies) plt.ylabel('Companies') plt.xlabel('Sales (USD)') plt.title('Company sales') plt.show()

#Sin Walmart fig = plt.figure(figsize=(10,10)) plt.grid(axis='both',zorder=0) sns.histplot(df_sorted[1:], x = 'Sales', y='Company',hue='Company',zorder=10) plt.legend('off') plt.xlabel('Sales (USD)')

#Comparativo acum = df_sorted[1:-16]['Sales'].sum() #Suman lo mismo que walmart walmart = df_sorted.iloc[0]['Sales'] resto = df_sorted[16:]['Sales'].sum() comp = [walmart,acum,resto] plt.bar(['Walmart','Top 15 Companies','Last Companies'],comp,color=['lightskyblue','crimson','yellowgreen']) plt.ylabel('Sales (USD)') plt.title('Company sales')

df_sorted

df_store_sorted = df_sorted.sort_values('Stores') df_store_sorted = df_store_sorted[df_store_sorted['Stores'].notna()] companies =df_store_sorted['Company'] stores = df_store_sorted['Stores'] fig = plt.figure(figsize = (10,10)) ax = fig.add_subplot() plt.stem(stores, orientation='horizontal',linefmt='k',markerfmt='C2o') ax.set_yticks([i for i in range(len(companies))]) ax.set_yticklabels(companies) plt.ylabel('Companies') plt.xlabel('Stores') plt.grid(axis='y') y_range = np.arange(1, len(df_store_sorted.index) + 1) for (_, row), y in zip(df_store_sorted.iterrows(), y_range): plt.annotate(int(row['Stores']), (row["Stores"]+300, y-1.1)) plt.xlim(-20, max(stores)+ 2000) plt.title('Company stores') plt.show()

#Las compañias que más venden fig = plt.figure(figsize=(7,7)) ax = fig.add_subplot() jp = sns.scatterplot(data=df_sorted,x='Stores', y ='Sales',hue='Company',s=100) plt.tight_layout() #plt.legend('off') plt.legend(loc='upper right', title='Companies',bbox_to_anchor=(1.4, 1)) """ Walmart is the top one in sales but doesn't have too much stores as others """

#Las compañias que más venden fig = plt.figure(figsize=(7,7)) ax = fig.add_subplot() jp = sns.regplot(data=df_sorted[1:],x='Stores', y ='Sales') plt.tight_layout() #plt.legend('off') plt.legend(loc='upper right', title='Companies',bbox_to_anchor=(1.4, 1))

#Las compañias que más venden fig = plt.figure(figsize=(7,7)) ax = fig.add_subplot() legends= [] for category in set(df_sorted[1:]['Category']): df_category = df_sorted[1:][df_sorted['Category']==category] if len(df_category)>2: legends.append(category) jp = sns.regplot(data=df_category,x='Stores', y ='Sales',label=category) plt.legend() plt.tight_layout() #plt.legend('off') plt.show()

companies = df_sorted['Company'] sales = df_sorted['Sales'] categories = set(df_sorted['Category'][df_sorted['Category'].notna()]) rango = sales.iloc[0]-sales.iloc[-1] print(FG_BLUE) print('El rango de ventas de todas las compañias: %i USD'%int(rango)) rango = sales.iloc[1]-sales.iloc[-1] print('El rango para todas las compañias (sin Walmart) is: %i USD'%int(rango)) print('--------------------------------') print(' Rango por categoría ') print('--------------------------------') for category in categories: df_category = df_sorted[df_sorted['Category']==category] if len(df_category)>1: print(FG_BLUE) rango = df_category['Sales'].iloc[0] - df_category['Sales'].iloc[-1] print('El rango por categoría %s is %i USD'%(category,int(rango))) else: print(FG_RED) print('La categoría %s tiene sólo una compañia'%category)

df_top_stores = df_sorted.sort_values('Stores')[df_sorted['Stores'].notna()][::-1][:5] df_top_stores

df_top_sales = df_sorted[:5] int_df = pd.merge(df_top_stores, df_top_sales, how ='inner', on =['Company', 'Sales','Stores']) int_df

df_sorted.groupby('Category').sum().sort_values('Sales',ascending=False).index[0]

df_sales = df_sorted['Sales'][:] df_stores = df_sorted['Stores'][:] df_sales_by_store = df_sales / df_stores df_sorted['Sales by store'] = df_sales_by_store df_sorted.sort_values('Sales by store', ascending = False)

sales_by_store = df_sorted.sort_values('Sales by store')[df_sorted['Sales by store'].notna()]['Sales by store'] companies = df_sorted.sort_values('Sales by store')[df_sorted['Sales by store'].notna()]['Company'] df_sorted_salesbystore = df_sorted.sort_values('Sales by store')[df_sorted['Sales by store'].notna()] fig = plt.figure(figsize = (10,10)) ax = fig.add_subplot() plt.stem(sales_by_store, orientation='horizontal',linefmt='k',markerfmt='C2o') ax.set_yticks([i for i in range(len(companies))]) ax.set_yticklabels(companies) plt.ylabel('Companies') plt.xlabel('Sales/Stores') plt.grid(axis='y') y_range = np.arange(1, len(df_sorted_salesbystore.index) + 1) for (_, row), y in zip(df_sorted_salesbystore.iterrows(), y_range): plt.annotate(int(row['Sales by store']), (row["Sales by store"]+10, y-1.1)) plt.xlim(-20, max(sales_by_store)+ 100) plt.title('Company ratio sales/stores') plt.show()

no_store_companies = df_sorted[df_sorted['Stores'].isna()]['Company'] no_store_companies

for comp in no_store_companies: print("La compañia %s esta en la posición n°%i en ventas"%( comp ,df_sorted.index[df_sorted['Company']==comp][0]))

df_category_means = df_sorted[df_sorted['Stores'].notna()].groupby('Category').mean().sort_values('Stores', ascending = False) df_category_means

stores_by_category = df_category_means['Stores'][::-1] category = df_category_means.index[::-1] fig = plt.figure(figsize = (10,10)) ax = fig.add_subplot() plt.stem(stores_by_category, orientation='horizontal',linefmt='k',markerfmt='C3o') ax.set_yticks([i for i in range(len(category))]) ax.set_yticklabels(category) plt.ylabel('Category') plt.xlabel('Av.Stores') plt.grid(axis='y') y_range = np.arange(1, len(category) + 1) for stor, y in zip(stores_by_category, y_range): plt.annotate(int(stor), (stor+300, y-1.1)) plt.xlim(-20, max(stores_by_category)+ 2000) plt.title('Av.Stores by category') plt.show()

df_company_growth = df_sorted.sort_values('Store Count Growth', ascending=False)[df_sorted['Store Count Growth'].notna()]

df_company_growth.iloc[0]

store_growth = df_company_growth['Store Count Growth'][::-1]*100 company = df_company_growth['Company'][::-1] fig = plt.figure(figsize = (10,10)) ax = fig.add_subplot() plt.stem(store_growth, orientation='horizontal',linefmt='k',markerfmt='C3o') ax.set_yticks([i for i in range(len(company))]) ax.set_yticklabels(company) plt.ylabel('Company') plt.xlabel('Store Growth (%)') plt.grid(axis='y') y_range = np.arange(1, len(company) + 1) for stor, y in zip(store_growth, y_range): x = stor +2 if stor >0 else stor-8 plt.annotate("{:10.2f}".format(stor), (x, y-1.1)) plt.xlim(-35, 35) plt.title('Store growth by company') plt.show()

growth_by_sales = store_growth/sales df_company_growth['Store Growth by Sales'] = growth_by_sales df_company_growth = df_company_growth.sort_values('Store Growth by Sales',ascending=False) sales = df_company_growth['Sales'][::1] growth_by_sales = df_company_growth['Store Growth by Sales'][::-1] company = df_company_growth['Company'][::-1] fig = plt.figure(figsize = (10,10)) ax = fig.add_subplot() plt.stem(growth_by_sales, orientation='horizontal',linefmt='k',markerfmt='C3o') ax.set_yticks([i for i in range(len(company))]) ax.set_yticklabels(company) plt.ylabel('Company') plt.xlabel('Store Growth by Sales (%)') plt.grid(axis='y') y_range = np.arange(1, len(company) + 1) for stor, y in zip(growth_by_sales, y_range): x = stor +0.0001 if stor >0 else stor-0.0002 plt.annotate("{:.2e}".format(stor), (x, y-1.1)) plt.xlim(-0.001,0.001) plt.title('Store growth by Sales company') plt.show()