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

df = pd.read_csv('/work/data/df_cleaned.csv') df

df.dtypes

df['fper'] = pd.to_datetime(df['fper']) df['despropie'] = df['despropie'].astype('category') df['destipbr'] = df['destipbr'].astype('category') df['descimi'] = df['descimi'].astype('category') df['desestru'] = df['desestru'].astype('category') df['despared'] = df['despared'].astype('category') df['descubi'] = df['descubi'].astype('category') df['descres'] = df['descres'].astype('category') df['DESCTIPOBR'] = df['DESCTIPOBR'].astype('category') df['DESCDISPSUT'] = df['DESCDISPSUT'].astype('category') df['DESCDISPUSO'] = df['DESCDISPUSO'].astype('category') df['DESCDISPARC'] = df['DESCDISPARC'].astype('category') df['DESCDISPRFI'] = df['DESCDISPRFI'].astype('category') df['DESCDISPRFX'] = df['DESCDISPRFX'].astype('category') df['DESCAREAUR'] = df['DESCAREAUR'].astype('category')

df.dtypes

df['DESCAREAUR'].unique()

sns.set_theme()

sns.countplot(y='desprovin', data=df)

df.groupby(['desprovin'])[['descodcantf']].count().sort_values(by='descodcantf', ascending=False)

df_5_provinces = df.groupby(['desprovin'])[['descodcantf']].count().sort_values(by='descodcantf', ascending=False).head(5).sum() df_rest_provinces = df.groupby(['desprovin'])[['descodcantf']].count().sort_values(by='descodcantf', ascending=False)[5:].sum() sizes_pie = [float(df_5_provinces), float(df_rest_provinces)] labels_pie = ['5 Provinces with Most Constructions', 'Rest of Provinces'] fig, ax = plt.subplots(figsize=(10, 7)) ax.pie(sizes_pie, labels=labels_pie, autopct='%1.1f%%') ax.axis('equal') plt.title('The 5 Provinces with most constructions compared to the rest') plt.show()

df.groupby(['descodcantf'])[['desmes']].count().sort_values(by='desmes', ascending=False)

df_5_cantons = df.groupby(['descodcantf'])[['desmes']].count().sort_values(by='desmes', ascending=False).head(5).sum() df_rest_cantons = df.groupby(['descodcantf'])[['desmes']].count().sort_values(by='desmes', ascending=False)[5:].sum() sizes_pie = [float(df_5_cantons), float(df_rest_cantons)] labels_pie = ['5 Cantons with Most Constructions', 'Rest of Cantons'] fig, ax = plt.subplots(figsize=(10, 7)) ax.pie(sizes_pie, labels=labels_pie, autopct='%1.1f%%') ax.axis('equal') plt.title('The 5 Cantons with most constructions compared to the rest') plt.show()

df.groupby(['desprovin'])[['CVAE']].sum().sort_values(by='CVAE', ascending=False)

df_5_provinces_cost = df.groupby(['desprovin'])[['CVAE']].sum().sort_values(by='CVAE', ascending=False).head(5).sum() df_rest_provinces_cost = df.groupby(['desprovin'])[['CVAE']].sum().sort_values(by='CVAE', ascending=False)[5:].sum() sizes_pie = [float(df_5_provinces_cost), float(df_rest_provinces_cost)] labels_pie = ['5 Provinces with Most Spending', 'Rest of Provinces'] fig, ax = plt.subplots(figsize=(10, 7)) ax.pie(sizes_pie, labels=labels_pie, autopct='%1.1f%%') ax.axis('equal') plt.title('The 5 Provinces with most construction spending compared to the rest') plt.show()

df.groupby(['descodcantf'])[['CVAE']].sum().sort_values(by='CVAE', ascending=False)

df_5_cantons_cost = df.groupby(['descodcantf'])[['CVAE']].sum().sort_values(by='CVAE', ascending=False).head(5).sum() df_rest_cantons_cost = df.groupby(['descodcantf'])[['CVAE']].sum().sort_values(by='CVAE', ascending=False)[5:].sum() sizes_pie = [float(df_5_cantons_cost), float(df_rest_cantons_cost)] labels_pie = ['5 Cantons with Most Spending', 'Rest of Cantons'] fig, ax = plt.subplots(figsize=(10, 7)) ax.pie(sizes_pie, labels=labels_pie, autopct='%1.1f%%') ax.axis('equal') plt.title('The 5 cantons with most construction spending compared to the rest') plt.show()

df_date_index = df.set_index('fper')

df_date_index['Quarter'] = df_date_index.index.quarter df_date_index['Month'] = df_date_index.index.month df_date_index['Day'] = df_date_index.index.weekday

df_date_index

df_date_index.groupby('Quarter')[['CVAE']].count().plot.bar(rot=0)

df_date_index.groupby('Quarter')[['CVAE']].sum().plot.bar(rot=0)

df_date_index.groupby('Month')[['CVAE']].count().plot.bar(rot=0)

df_date_index.groupby('Month')[['CVAE']].sum().plot.bar(rot=0)

sns.countplot(y='despropie', data=df_date_index)

df_date_index.groupby('despropie')[['CVAE']].sum().plot.bar(rot=0)

pd.options.display.float_format = '{:,.2f}'.format

df_date_index.groupby(['despropie', 'DESCDISPUSO'])[['CVAE']].median()

sns.countplot(y='destipbr', data=df_date_index)

sns.countplot(y='descimi', data=df_date_index)

sns.countplot(y='desestru', data=df_date_index)

sns.countplot(y='despared', data=df_date_index)

sns.countplot(y='descubi', data=df_date_index)

sns.countplot(y='descres', data=df_date_index)

sns.countplot(y='DESCTIPOBR', data=df_date_index)

sns.countplot(y='DESCDISPSUT', data=df_date_index)

sns.countplot(y='DESCDISPUSO', data=df_date_index)

sns.countplot(y='DESCDISPARC', data=df_date_index)

sns.countplot(data=df_date_index, y='DESCDISPRFI')

sns.countplot(y='DESCDISPRFX', data=df_date_index)

sns.countplot(y='DESCAREAUR', data=df_date_index)

sns.displot(data=df_date_index, x='nucu', kde=True)

df_date_index[['nucu']].describe()

sns.displot(data=df_date_index, x='nudo', kde=True)

df_date_index[['nudo']].describe()

sns.displot(data=df_date_index, x='nupi', kde=True)

df_date_index[['nupi']].describe()

df_date_index[['nupi', 'tonp', 'NUVICAL']].describe()

df_date_index.describe()

#Type of property (private/public) and type of construction sns.countplot(x='despropie', hue='destipbr' ,data=df_date_index)

#Public property and type of construction sns.countplot(data=df_date_index[(df_date_index['despropie'] == 'Pública')], x='despropie', hue='destipbr') plt.legend(loc='upper left')

#Type of property (private/public) and type of structure sns.countplot(x='despropie', hue='desestru' ,data=df_date_index) plt.legend(loc='upper right')

#Public property and type of construction sns.countplot(data=df_date_index[(df_date_index['despropie'] == 'Pública')], x='despropie', hue='desestru') plt.legend(loc='upper left')

#Type of property (private/public) and type of use sns.countplot(hue='despropie', y='DESCDISPUSO' ,data=df_date_index) plt.legend(loc='upper right')

#Public property and type of construction sns.countplot(data=df_date_index[(df_date_index['despropie'] == 'Pública')], y='DESCDISPUSO', hue='despropie') plt.legend(loc='lower right')

#Type of property (private/public) and financing type sns.countplot(hue='despropie', y='DESCDISPRFI' ,data=df_date_index) plt.legend(loc='upper right')

#Public property and financing type sns.countplot(data=df_date_index[(df_date_index['despropie'] == 'Pública')], y='DESCDISPRFI', hue='despropie') plt.legend(loc='lower right')

#Quarter df_date_index.groupby(['Quarter', 'despropie'])[['despropie']].count()

df_date_index[(df_date_index['despropie'] == 'Pública')].groupby(['Quarter', 'despropie'])[['desestru']].count().plot.bar(rot=45)

df_price = df_date_index.groupby(['despropie', 'DESCDISPUSO'])[['CVAE', 'CARCO']].sum()

df_price['Price_m2'] = df_price['CVAE'] / df_price['CARCO']

df_price

df_price['Price_m2'] = df_price['Price_m2'].fillna(0)

fig, ax = plt.subplots(figsize=(10, 7)) sns.lineplot(data=df_price, x='DESCDISPUSO', y='Price_m2', hue='despropie') plt.xticks(rotation=90)

fig, ax = plt.subplots(figsize=(10, 7)) sns.barplot(data=df_price.reset_index(), x='DESCDISPUSO', y='Price_m2', hue='despropie') plt.xticks(rotation=90)

df_date_index['Price_m2'] = df_date_index['CVAE'] / df_date_index['CARCO']

df_date_index['Price_m2'] = df_date_index['Price_m2'].fillna(0)

df_date_index['Price_m2'].describe()

fig, ax = plt.subplots(figsize=(40, 40)) sns.heatmap(df_date_index.corr(), annot=True, square=True, annot_kws={"fontsize":15})

df_date_index['DESCDISPRFI'].unique()

df_date_index.dtypes

#Correlation number of floors and rooms sns.regplot(data=df_date_index, x='tonp', y='NUCUCAL') plt.ylim(0, 50)

#Corr construction area and green area sns.regplot(data=df_date_index, x='CARCO', y='CAESV')

#Corr land area and construction area sns.regplot(data=df_date_index, x='CSUTE', y='CARCO')

#Corr land area and green area sns.regplot(data=df_date_index, x='CSUTE', y='CAESV')

#Corr construction area and green area sns.regplot(data=df_date_index, x='CARCO', y='CAESV')

#Corr construction area and total price of construction sns.regplot(data=df_date_index, x='CARCO', y='CVAE')

#Corr land and total price of construction sns.regplot(data=df_date_index, x='CSUTE', y='CVAE')

#Corr green area and total price of construction sns.regplot(data=df_date_index, x='CAESV', y='CVAE')

df_date_index.dtypes

#Correlation number of floors and rooms sns.scatterplot(data=df_date_index, x='tonp', y='NUCUCAL', hue='despropie') plt.ylim(0, 50)

#Corr construction area and green area by type of property sns.scatterplot(data=df_date_index, x='CARCO', y='CAESV', hue='despropie')

#Corr land area and construction area by type of property sns.scatterplot(data=df_date_index, x='CSUTE', y='CARCO', hue='despropie')

#Corr land area and green area sns.scatterplot(data=df_date_index, x='CSUTE', y='CAESV', hue='despropie')

df_date_index.groupby(['despropie', 'desestru'])[['tonp']].max()

#Corr construction area and total price of construction by type of property sns.scatterplot(data=df_date_index, x='CARCO', y='CVAE', hue='despropie') plt.ylim(0, 2000000) plt.xlim(0, 15000)

#Corr green area and total price of construction by type of property sns.scatterplot(data=df_date_index, x='CAESV', y='CVAE', hue='despropie') plt.ylim(0, 2000000) plt.xlim(0, 15000)

#Corr construction area and total price of construction by type of structure sns.scatterplot(data=df_date_index, x='CARCO', y='CVAE', hue='desestru')

#Corr green area and total price of construction by type of structure sns.scatterplot(data=df_date_index, x='CAESV', y='CVAE', hue='desestru')

#Corr construction area and green area by type of construction sns.scatterplot(data=df_date_index, x='CARCO', y='CAESV', hue='desestru')