|
import streamlit as st |
|
import hopsworks |
|
import joblib |
|
import pandas as pd |
|
|
|
import numpy as np |
|
import folium |
|
from streamlit_folium import st_folium, folium_static |
|
import json |
|
import time |
|
from datetime import timedelta, datetime |
|
from branca.element import Figure |
|
|
|
from functions import decode_features, get_model |
|
|
|
|
|
def fancy_header(text, font_size=24): |
|
res = f'<span style="color:#ff5f27; font-size: {font_size}px;">{text}</span>' |
|
st.markdown(res, unsafe_allow_html=True ) |
|
|
|
|
|
st.title('⛅️Air Quality Prediction Project🌩') |
|
|
|
progress_bar = st.sidebar.header('⚙️ Working Progress') |
|
progress_bar = st.sidebar.progress(0) |
|
st.write(36 * "-") |
|
fancy_header('\n📡 Connecting to Hopsworks Feature Store...') |
|
|
|
project = hopsworks.login(api_key_value="0rdWXlLgEd3mkGOg.iRZ7TtAkWGPlJHNQcAEph6Qbokoaq7QTBRI9ckwWUki8tIYGyBvrKhJvtLoUOGQ4") |
|
fs = project.get_feature_store() |
|
feature_view = fs.get_feature_view( |
|
name = 'air_quality_fv', |
|
version = 1 |
|
) |
|
|
|
st.write("Successfully connected!✔️") |
|
progress_bar.progress(20) |
|
|
|
st.write(36 * "-") |
|
fancy_header('\n☁️ Getting batch data from Feature Store...') |
|
|
|
start_date = datetime.now() - timedelta(days=1) |
|
start_time = int(start_date.timestamp()) * 1000 |
|
|
|
X = feature_view.get_batch_data(start_time=start_time) |
|
progress_bar.progress(50) |
|
|
|
latest_date_unix = str(X.date.values[0])[:10] |
|
latest_date = time.ctime(int(latest_date_unix)) |
|
|
|
st.write(f"⏱ Data for {latest_date}") |
|
|
|
X = X.drop(columns=["date"]).fillna(0) |
|
|
|
data_to_display = decode_features(X, feature_view=feature_view) |
|
|
|
progress_bar.progress(60) |
|
|
|
st.write(36 * "-") |
|
fancy_header(f"🗺 Processing the map...") |
|
|
|
fig = Figure(width=550,height=350) |
|
|
|
my_map = folium.Map(location=[58, 20], zoom_start=3.71) |
|
fig.add_child(my_map) |
|
folium.TileLayer('Stamen Terrain').add_to(my_map) |
|
folium.TileLayer('Stamen Toner').add_to(my_map) |
|
folium.TileLayer('Stamen Water Color').add_to(my_map) |
|
folium.TileLayer('cartodbpositron').add_to(my_map) |
|
folium.TileLayer('cartodbdark_matter').add_to(my_map) |
|
folium.LayerControl().add_to(my_map) |
|
|
|
data_to_display = data_to_display[["city", "temp", "humidity", |
|
"conditions", "aqi"]] |
|
|
|
cities_coords = {("Sundsvall", "Sweden"): [62.390811, 17.306927], |
|
("Stockholm", "Sweden"): [59.334591, 18.063240], |
|
("Malmo", "Sweden"): [55.604981, 13.003822]} |
|
|
|
if "Kyiv" in data_to_display["city"]: |
|
cities_coords[("Kyiv", "Ukraine")]: [50.450001, 30.523333] |
|
|
|
data_to_display = data_to_display.set_index("city") |
|
|
|
cols_names_dict = {"temp": "Temperature", |
|
"humidity": "Humidity", |
|
"conditions": "Conditions", |
|
"aqi": "AQI"} |
|
|
|
data_to_display = data_to_display.rename(columns=cols_names_dict) |
|
|
|
cols_ = ["Temperature", "Humidity", "AQI"] |
|
data_to_display[cols_] = data_to_display[cols_].apply(lambda x: round(x, 1)) |
|
|
|
for city, country in cities_coords: |
|
text = f""" |
|
<h4 style="color:green;">{city}</h4> |
|
<h5 style="color":"green"> |
|
<table style="text-align: right;"> |
|
<tr> |
|
<th>Country:</th> |
|
<td><b>{country}</b></td> |
|
</tr> |
|
""" |
|
for column in data_to_display.columns: |
|
text += f""" |
|
<tr> |
|
<th>{column}:</th> |
|
<td>{data_to_display.loc[city][column]}</td> |
|
</tr>""" |
|
text += """</table> |
|
</h5>""" |
|
|
|
folium.Marker( |
|
cities_coords[(city, country)], popup=text, tooltip=f"<strong>{city}</strong>" |
|
).add_to(my_map) |
|
|
|
|
|
|
|
folium_static(my_map) |
|
progress_bar.progress(80) |
|
st.sidebar.write("-" * 36) |
|
|
|
|
|
model = get_model(project=project, |
|
model_name="gradient_boost_model", |
|
evaluation_metric="f1_score", |
|
sort_metrics_by="max") |
|
|
|
preds = model.predict(X) |
|
|
|
cities = [city_tuple[0] for city_tuple in cities_coords.keys()] |
|
|
|
next_day_date = datetime.today() + timedelta(days=1) |
|
next_day = next_day_date.strftime ('%d/%m/%Y') |
|
df = pd.DataFrame(data=preds, index=cities, columns=[f"AQI Predictions for {next_day}"], dtype=int) |
|
|
|
st.sidebar.write(df) |
|
progress_bar.progress(100) |
|
st.button("Re-run") |