import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt
import numpy as np
import scipy as sp

# 1
df = pd.read_csv('epa-sea-level.csv')
df

	Year	CSIRO Adjusted Sea Level	Lower Error Bound	Upper Error Bound	NOAA Adjusted Sea Level
0	1880	0.000000	-0.952756	0.952756	NaN
1	1881	0.220472	-0.732283	1.173228	NaN
2	1882	-0.440945	-1.346457	0.464567	NaN
3	1883	-0.232283	-1.129921	0.665354	NaN
4	1884	0.590551	-0.283465	1.464567	NaN
...	...	...	...	...	...
129	2009	8.586614	8.311024	8.862205	8.046354
130	2010	8.901575	8.618110	9.185039	8.122973
131	2011	8.964567	8.661417	9.267717	8.053065
132	2012	9.326772	8.992126	9.661417	8.457058
133	2013	8.980315	8.622047	9.338583	8.546648

134 rows × 5 columns

# 2
def draw_sea_level_data(df: pd.DataFrame = df):
    fig, ax = plt.subplots(figsize = (8, 5))
    sns.scatterplot(data=df, x = 'Year', y = 'CSIRO Adjusted Sea Level', ax=ax)
    ax.set_xlabel('Year')
    ax.set_ylabel('Sea Level (inches)')
    ax.set_title('Rise in Sea Level')
    ax.grid(True, ls = ':')

    plt.tight_layout()
    plt.savefig('SIRO_Adjusted_Sea_Level_scatter_plot.png')

draw_sea_level_data(df=df)

def reg(df: pd.DataFrame = df):
    pred_df = df.copy()
        
    slope, intercept, r, p, se = sp.stats.linregress(pred_df['Year'], pred_df['CSIRO Adjusted Sea Level'])
    return slope, intercept

# 3.1
def sea_level_reg(df: pd.DataFrame = df):
    df = df.copy()
    
    slope, intercept = reg(df=df)
    
    fig, ax = plt.subplots(figsize = (8, 5))
    sns.scatterplot(data=df, x = 'Year', y = 'CSIRO Adjusted Sea Level', ax=ax, label = 'original data')
    ax.plot(df['Year'], intercept + slope*df['Year'], c = 'r', label = 'fitted line')
    ax.set_xlabel('Year')
    ax.set_ylabel('Sea Level (inches)')
    ax.set_title('Rise in Sea Level')
    ax.grid(True, ls = ':')
    ax.legend()

    plt.tight_layout()
    plt.savefig('SIRO_Adjusted_Sea_Level_regression_curve.png')

sea_level_reg(df=df)

# 3.2
def sea_level_pred_overall(df: pd.DataFrame = df, pred_to_date: int = None):
    df = df.copy()
    extended_time = np.arange(df['Year'].min(), pred_to_date)
    
    slope, intercept = reg(df=df)
    
    fig, ax = plt.subplots(figsize = (8, 5))
    sns.scatterplot(data=df, x = 'Year', y = 'CSIRO Adjusted Sea Level', ax=ax, label = 'original data')
    ax.plot(extended_time, intercept + slope*extended_time, c = 'r', label = 'fitted line')
    ax.set_xlabel('Year')
    ax.set_ylabel('Sea Level (inches)')
    ax.set_title('Rise in Sea Level')
    ax.grid(True, ls = ':')
    ax.legend()

    plt.tight_layout()
    plt.savefig('SIRO_Adjusted_Sea_Level_regression_curve_with_prediction_overall.png')

sea_level_pred_overall(df=df, pred_to_date=2050)

# 4.1
def sea_level_reg_from_current_year(df: pd.DataFrame = df, current_year: int = None):
    df = df.copy()
    df = df[df['Year'] >= current_year]
    
    slope, intercept = reg(df=df)
    
    fig, ax = plt.subplots(figsize = (8, 5))
    sns.scatterplot(data=df, x = 'Year', y = 'CSIRO Adjusted Sea Level', ax=ax, label = 'original data')
    ax.plot(df['Year'], intercept + slope*df['Year'], c = 'r', label = 'fitted line')
    ax.set_xlabel('Year')
    ax.set_ylabel('Sea Level (inches)')
    ax.set_title('Rise in Sea Level')
    ax.grid(True, ls = ':')
    ax.legend()

    plt.tight_layout()
    plt.savefig('SIRO_Adjusted_Sea_Level_regression_curve_from_current_year.png')
sea_level_reg_from_current_year(df=df, current_year=2000)

# 4.2
def sea_level_pred_from_current_year_to_date(df: pd.DataFrame = df, current_year: int = None, pred_to_date: int = None):
    df = df.copy()
    df = df[df['Year'] >= current_year]
    extended_time = np.arange(df['Year'].min(), pred_to_date)
    
    slope, intercept = reg(df=df)
    
    fig, ax = plt.subplots(figsize = (8, 5))
    sns.scatterplot(data=df, x = 'Year', y = 'CSIRO Adjusted Sea Level', ax=ax, label = 'original data')
    ax.plot(extended_time, intercept + slope*extended_time, c = 'r', label = 'fitted line')
    ax.set_xlabel('Year')
    ax.set_ylabel('Sea Level (inches)')
    ax.set_title('Rise in Sea Level')
    ax.grid(True, ls = ':')
    ax.legend()

    plt.tight_layout()
    plt.savefig('SIRO_Adjusted_Sea_Level_regression_curve_from_current_year_to_date.png')
    
sea_level_pred_from_current_year_to_date(df=df, current_year=2000, pred_to_date=2050)