FoodQast / visualise.py
Silverxwolf's picture
Upload 12 files
ca8bb53
Raw History Blame Contribute Delete
4.15 kB
import streamlit as st
import matplotlib.pyplot as plt
from math import sqrt
import numpy as np
import pandas as pd
class DataVisualizer:
def __init__(self, datasets, load_options):
self.datasets = datasets
self.load_options = load_options
self.plots = dict()
self.tables = dict()
def visualise_data(self):
fig_main, ax_main = plt.subplots(1, 1, figsize=(15, 5))
if self.load_options["evaluate"]:
tmp = self.datasets["train_df"].append(self.datasets["test_df"].iloc[0], ignore_index=True)
ax_main.plot(tmp["date"], tmp["food_amount"])
ax_main.plot(self.datasets["test_df"]["date"], self.datasets["test_df"]["food_amount"])
else:
ax_main.plot(self.datasets["food_df"]["date"], self.datasets["food_df"]["food_amount"])
self.plots["visualise_data"] = fig_main
st.write(fig_main)
def visualise_evaluation(self, results):
tab1, tab2 = st.tabs(["Graph", "Table"])
with tab1:
x = np.arange(len(results["evaluation_real"]))
f, ax = plt.subplots(1)
f.set_figheight(5)
f.set_figwidth(15)
ax.plot(x, results["evaluation_real"], marker='o', markersize=10, linestyle='', color='black', label='Actual data')
ax.plot(x, results["evaluation_results"], marker='o', markersize=10, linestyle='', color='orange', label='Forecast')
ax.grid()
plt.xlabel('Time', fontsize=16)
plt.ylabel(self.load_options["frequency"] + ' Demand', fontsize=16)
ax.legend(loc='best')
plt.suptitle(self.load_options["frequency"] + ' forecast', y=1)
ax.set_xticks(x)
ax.set_xticklabels(x.astype(int))
self.plots["evaluation_graph"] = f
st.write(f)
with tab2:
test_dates = self.datasets["test_df"]["date"]
evaluation_results = results["evaluation_results"]
evaluation_table = pd.DataFrame({'Date': test_dates, 'Evaluation_Result': evaluation_results})
self.tables["evaluation_table"] = evaluation_table
st.dataframe(evaluation_table, use_container_width=True)
def visualise_forecast(self, results):
tab1, tab2 = st.tabs(["Graph", "Table"])
forecast_results = results["forecast_results"]
with tab1:
x = np.arange(len(forecast_results))
f, ax = plt.subplots(1)
f.set_figheight(5)
f.set_figwidth(15)
ax.plot(x, forecast_results, marker='o', markersize=10, linestyle='', color='orange', label='Forecast')
ax.grid()
plt.xlabel('Time', fontsize=16)
plt.ylabel(self.load_options["frequency"] + ' Demand', fontsize=16)
ax.legend(loc='best')
plt.suptitle(self.load_options["frequency"] + ' forecast', y=1)
ax.set_xticks(x)
ax.set_xticklabels(x.astype(int))
self.plots["visualise_forecast"] = f
st.write(f)
with tab2:
start_date = self.datasets["forecast_df"]["date"].max()
date_offset = pd.DateOffset(**{self.load_options["freq_label"]: 1})
date_range = pd.date_range(start_date + date_offset,
periods=self.load_options["duration"],
freq= self.load_options["freq"]
)
forecast_table = pd.DataFrame({'Date': date_range, 'Forecast_Result': forecast_results})
self.tables["forecast_table"] = forecast_table
st.dataframe(forecast_table, use_container_width=True)
def display_evaluation(self, results):
eval = results["evaluation_metrics"]
evaluation = pd.DataFrame(
[[eval["MAE"], eval["MAPE"], sqrt(eval["MSE"])]],
columns = ["MAE", "MAPE", "RMSE"]
)
self.tables["evaluation_metrics"] = evaluation
st.table(evaluation)