app.py

# -*- coding: utf-8 -*-
"""
Created on Thu Jun  1 13:52:42 2023

@author: ME
"""

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import streamlit as st
import seaborn as sns
import pickle
import base64
import joblib
from src.preprocess import preprocess_data
import mplcursors


#load json file
json_path = r"C:/Users/ME/Desktop/Blessing_AI/Weather_Prediction/Artifacts/feature_dict.joblib"
loaded_data = joblib.load(json_path)

#load model
model_path = r"C:/Users/ME/Desktop/Blessing_AI/Weather_Prediction/Artifacts/trained_prophet_model.pkl"
with open(model_path, 'rb') as file_:
    model = pickle.load(file_)
    

st.title("Weather Forecast: Average Daily Temperature Projections for Nigerian Airports")

col1,col2 = st.columns(2)

start_d = col1.date_input("Select start date")
end_d =  col2.date_input("Select end date")
#get number of days
delta = end_d - start_d
days = delta.days

#split start date into year ,month and date 
y1 = start_d.year
m1 = start_d.month
d1  = start_d.day

#split end date into year ,month and date
y2 = end_d.year
m2 = end_d.month
d2  = end_d.day

start_date = (y1,m1,d1)
end_date = (y2,m2,d2)

if days > 10 :
    st.write("Select a lower number of days range (10 and below)")
    
else:
    
    name = st.selectbox("Select airport of interest",options=tuple(loaded_data["Airport_name"].keys()))

    data = preprocess_data(
        start_d=start_date,
        end_d= end_date,
       airport_name = name
        )
    
    forecast_bn = st.button("Forecast")
    
    if forecast_bn:
        pred_df = model.predict(data)
        pred_df = pred_df[["ds","yhat"]]
        pred_df.columns = ["Date","Prediction"]
        
        #Display only year,month and date
        pred_df["Date"]= pred_df['Date'].dt.strftime('%Y-%m-%d')
        
        #Convert column to 2 dp
        pred_df["Prediction"] = pred_df["Prediction"].round(2)
        
        tab1,tab2 = st.tabs(["Predictions","Check plot"])
        tab1.write(pred_df)
        
        # Add a download button
        csv = pred_df.to_csv(index=False)
        b64 = base64.b64encode(csv.encode()).decode()  # Encode the DataFrame as a base64 string
        href = f'<a href="data:file/csv;base64,{b64}" download="data.csv">Download CSV File</a>'
        st.markdown(href, unsafe_allow_html=True)
                
        
        # Create the figure and axis objects
        fig, ax = plt.subplots()
        
        # Plot the time series
        x = pred_df["Date"]
        y = pred_df["Prediction"]
        ax.plot(x, y, marker='o', linestyle='--')
        
        # Set x-axis label
        ax.set_xlabel('Date')
        
        # Set y-axis label
        ax.set_ylabel('Average temperature')
        
        # Style the plot
        sns.set_style('whitegrid')
        
        # Set the limits of the x-axis and y-axis
        ax.set_xlim(min(x), max(x))
        ax.set_ylim(min(y), max(y))
        
        # Format the x-axis date ticks
        fig.autofmt_xdate()
        
        # Add cursor interaction
        cursor = mplcursors.cursor(hover=True)
        # Define the annotation format
        @cursor.connect("add")
        def on_add(sel):
            x = sel.target[0]
            y = sel.target[1]
            date_str = x.strftime("%Y-%m-%d")
            sel.annotation.set_text(f'Date: {date_str}\nValue: {y:.2f}')
        
        # Show the plot
        tab2.pyplot(fig)