!pip install statsmodels==0.14.0

import pandas as pd import seaborn as sns import matplotlib.pyplot as plt import matplotlib.dates as mdates import numpy as np import scipy.stats as stats import math from sklearn.model_selection import train_test_split from sklearn.linear_model import LinearRegression from sklearn.metrics import mean_squared_error, mean_absolute_error from sklearn.metrics import mean_absolute_percentage_error from sklearn.metrics import median_absolute_error from sklearn.preprocessing import MinMaxScaler, StandardScaler

df = pd.read_csv("datasets/AT_BostonHousing_Data.csv", sep=";") df.head(10)

df.isna().sum()

df_numeric_stats = df[["age", "tax", "medv"]] df_numeric_stats.describe()

plt.figure(figsize=(12, 6)) plt.subplot(1, 3, 1) sns.boxplot(x=df_numeric_stats['age'], data=df) plt.title('Age Box Plot') plt.subplot(1, 3, 2) sns.boxplot(x=df_numeric_stats['tax'], data=df) plt.title('Tax Boxplot') plt.subplot(1, 3, 3) sns.boxplot(x=df_numeric_stats['medv'], data=df) plt.title('Medv Boxplot') plt.tight_layout() plt.show()

df['rm'].describe()

def plot_hist(df, col): plt.figure(figsize=(10, 6)) sns.histplot(df[col], bins=30, kde=True, color='red', edgecolor='black') plt.title(f'Distribuição da Coluna "{col}"') plt.xlabel(col) plt.ylabel('Frequência') plt.show() plot_hist(df, 'rm')

def plot_corr_heatmap(df): plt.figure(figsize=(12, 10)) correlation_matrix = df.corr() sns.heatmap(correlation_matrix, annot=True, cmap='coolwarm', fmt=".2f", linewidths=.5, mask=np.triu(np.ones_like(correlation_matrix, dtype=bool))) plt.title('Mapa de Calor - Correlação entre todas as variáveis') plt.show() plot_corr_heatmap(df)

dependent_var = 'medv' data_Y = df[dependent_var] data_X = df.drop(columns=[dependent_var]) print(f"Variáveis Independentes: {', '.join(data_X.columns)}") print(f"Variável Dependente: {dependent_var}")

g = sns.PairGrid(df, corner=True) g = g.map_lower(sns.regplot, scatter_kws={'alpha':0.15}, line_kws={'color': 'red'}) plt.show() # Salvar imagem ou abrir o link em outra aba, para melhor visualização

data_X = df[['zn', 'dis', 'rm']] data_X.head()

train_X, test_X, train_Y, test_Y = train_test_split(data_X, data_Y, test_size=0.25, shuffle=False) print(f"Train X: {train_X.shape} | Test X: {test_X.shape} | Train Y: {train_Y.shape[0]} | Test Y: {test_Y.shape[0]}")

def print_min_max(df, column): min_value = df[column].min() max_value = df[column].max() print(f"Column: {column}") print(f"Minimum Value: {min_value}") print(f"Maximum Value: {max_value}") print() column_name = 'your_column' print_min_max(df, 'zn') print_min_max(df, 'dis') print_min_max(df, 'rm')

MinMax_scaler = MinMaxScaler() train_X_minmax = MinMax_scaler.fit_transform(train_X) train_X_minmax

df_X_train_minmax = pd.DataFrame(train_X_minmax, columns=['zn','dis', 'rm']) df_X_train_minmax.describe()

regressor_simple = LinearRegression() train_X_rm = train_X['rm'].values.reshape(-1, 1) test_X_rm = test_X['rm'].values.reshape(-1, 1) regressor_simple.fit(train_X_rm, train_Y) predict_y_simple = regressor_simple.predict(test_X_rm) print('RMSE para regressão linear simples: ', math.sqrt(mean_squared_error(test_Y, predict_y_simple)))

regressor_multiple = LinearRegression() regressor_multiple.fit(train_X, train_Y) predict_y_multiple = regressor_multiple.predict(test_X) print('RMSE para regressão linear múltipla: ', math.sqrt(mean_squared_error(test_Y, predict_y_multiple)))

df_assets = pd.read_csv('datasets/AT_Ativos_PETR_VALE.csv', index_col='Date') df_assets.head()

df_assets['Date'] = df_assets.index df_long_assets = pd.melt(df_assets, id_vars='Date', var_name='Asset', value_name='Price') df_long_assets.index = df_long_assets['Date'] df_long_assets['Date'] = pd.to_datetime(df_long_assets['Date']) df_long_assets.head()

plt.figure(figsize=(20,10)) ax = sns.lineplot(df_long_assets, x='Date', y='Price', hue='Asset') plt.ylabel('Price', fontsize=10) plt.xticks(fontsize=9) loc = mdates.MonthLocator(interval=3) ax.xaxis.set_major_locator(loc) fmt = mdates.DateFormatter('%b\n%Y') ax.xaxis.set_major_formatter(fmt)

df_long_assets['Asset'].unique()

df_long_petr3 = df_long_assets[df_long_assets['Asset'] == 'PETR3'] df_long_petr3['Asset'].unique()

fig, ax = plt.subplots(figsize = (20, 8)) ax.plot(df_long_petr3['Date'], df_long_petr3['Price'], label = 'PETR3') ax.plot(df_long_petr3['Date'], df_long_petr3['Price'].rolling(window=15).mean(), label='Mean 15d') ax.plot(df_long_petr3['Date'], df_long_petr3['Price'].rolling(window=50).mean(), label='Mean 50d') ax.plot(df_long_petr3['Date'], df_long_petr3['Price'].rolling(window=100).mean(), label='Mean 100d') loc = mdates.MonthLocator(interval=3) ax.xaxis.set_major_locator(loc) fmt = mdates.DateFormatter('%b\n%Y') ax.xaxis.set_major_formatter(fmt) ax.legend() plt.show()

df_long_vale3 = df_long_assets[df_long_assets['Asset'] == 'VALE3'] df_long_vale3['Asset'].unique()

fig, ax = plt.subplots(figsize = (20, 8)) ax.plot(df_long_vale3['Date'], df_long_vale3['Price'], label = 'PETR3') ax.plot(df_long_vale3['Date'], df_long_vale3['Price'].rolling(window=200).mean(), label='Mean 200d') loc = mdates.MonthLocator(interval=3) ax.xaxis.set_major_locator(loc) fmt = mdates.DateFormatter('%b\n%Y') ax.xaxis.set_major_formatter(fmt) ax.legend() plt.show()