File size: 8,831 Bytes
9938325 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 |
import streamlit as st
import pandas as pd
import plotly.express as px
import plotly.graph_objects as go
import numpy as np
import pickle
import statsmodels.api as sm
import numpy as np
from sklearn.metrics import mean_absolute_error, r2_score,mean_absolute_percentage_error
from sklearn.preprocessing import MinMaxScaler
import matplotlib.pyplot as plt
from statsmodels.stats.outliers_influence import variance_inflation_factor
from plotly.subplots import make_subplots
st.set_option('deprecation.showPyplotGlobalUse', False)
from datetime import datetime
import seaborn as sns
def calculate_discount(promo_price_series, non_promo_price_series):
# Calculate the 4-week moving average of non-promo price
window_size = 4
base_price = non_promo_price_series.rolling(window=window_size).mean()
# Calculate discount_raw
discount_raw_series = (1 - promo_price_series / base_price) * 100
# Calculate discount_final
discount_final_series = discount_raw_series.where(discount_raw_series >= 5, 0)
return base_price, discount_raw_series, discount_final_series
def create_dual_axis_line_chart(date_series, promo_price_series, non_promo_price_series, base_price_series, discount_series):
# Create traces for the primary axis (price vars)
trace1 = go.Scatter(
x=date_series,
y=promo_price_series,
name='Promo Price',
yaxis='y1'
)
trace2 = go.Scatter(
x=date_series,
y=non_promo_price_series,
name='Non-Promo Price',
yaxis='y1'
)
trace3 = go.Scatter(
x=date_series,
y=base_price_series,
name='Base Price',
yaxis='y1'
)
# Create a trace for the secondary axis (discount)
trace4 = go.Scatter(
x=date_series,
y=discount_series,
name='Discount',
yaxis='y2'
)
# Create the layout with dual axes
layout = go.Layout(
title='Price and Discount Over Time',
yaxis=dict(
title='Price',
side='left'
),
yaxis2=dict(
title='Discount',
side='right',
overlaying='y',
showgrid=False
),
xaxis=dict(title='Date'),
)
# Create the figure with the defined traces and layout
fig = go.Figure(data=[trace1, trace2, trace3, trace4], layout=layout)
return fig
def to_percentage(value):
return f'{value * 100:.1f}%'
def plot_actual_vs_predicted(date, y, predicted_values, model,target_column=None, flag=None, repeat_all_years=False, is_panel=False):
if flag is not None :
fig = make_subplots(specs=[[{"secondary_y": True}]])
else :
fig = go.Figure()
if is_panel :
df=pd.DataFrame()
df['date'] = date
df['Actual'] = y
df['Predicted'] = predicted_values
df_agg = df.groupby('date').agg({'Actual':'sum', 'Predicted':'sum'}).reset_index()
df_agg.columns = ['date', 'Actual', 'Predicted']
assert len(df_agg) == pd.Series(date).nunique()
# date = df_agg['date']
# y = df_agg['Actual']
# predicted_values = df_agg['Predicted']
# ymax = df_agg['Actual'].max() # Sprint3 - ymax to set y value for flag
fig.add_trace(go.Scatter(x=df_agg['date'], y=df_agg['Actual'], mode='lines', name='Actual', line=dict(color='#08083B')))
fig.add_trace(go.Scatter(x=df_agg['date'], y=df_agg['Predicted'], mode='lines', name='Predicted', line=dict(color='#11B6BD')))
else :
fig.add_trace(go.Scatter(x=date, y=y, mode='lines', name='Actual', line=dict(color='#08083B')))
fig.add_trace(go.Scatter(x=date, y=predicted_values, mode='lines', name='Predicted', line=dict(color='#11B6BD')))
line_values=[]
if flag:
min_date, max_date = flag[0], flag[1]
min_week = datetime.strptime(str(min_date), "%Y-%m-%d").strftime("%U")
max_week = datetime.strptime(str(max_date), "%Y-%m-%d").strftime("%U")
month=pd.to_datetime(min_date).month
day=pd.to_datetime(min_date).day
#st.write(pd.to_datetime(min_date).week)
#st.write(min_week)
# Initialize an empty list to store line values
# Sprint3 change : put flags to secondary axis, & made their y value to 1 instead of 5M
if repeat_all_years:
#line_values=list(pd.to_datetime((pd.Series(date)).dt.week).map(lambda x: 10000 if x==min_week else 0 ))
#st.write(pd.Series(date).map(lambda x: pd.Timestamp(x).week))
line_values=list(pd.Series(date).map(lambda x: 1 if (pd.Timestamp(x).week >=int(min_week)) & (pd.Timestamp(x).week <=int(max_week)) else 0))
assert len(line_values) == len(date)
#st.write(line_values)
fig.add_trace(go.Scatter(x=date, y=line_values, mode='lines', name='Flag', line=dict(color='#FF5733')),secondary_y=True)
else:
line_values = []
line_values = list(pd.Series(date).map(lambda x: 1 if (pd.Timestamp(x) >= pd.Timestamp(min_date)) and (pd.Timestamp(x) <= pd.Timestamp(max_date)) else 0))
#st.write(line_values)
fig.add_trace(go.Scatter(x=date, y=line_values, mode='lines', name='Flag', line=dict(color='#FF5733')),secondary_y=True)
# Calculate MAPE
mape = mean_absolute_percentage_error(y, predicted_values)
print('mape*********',mape)
# Calculate AdjR2 # Assuming X is your feature matrix
r2 = r2_score(y, predicted_values)
adjr2 = 1 - (1 - r2) * (len(y) - 1) / (len(y) - len(model.params) - 1) #manoj
# Create a table to display the metrics
metrics_table = pd.DataFrame({
'Metric': ['MAPE', 'R-squared', 'AdjR-squared'],
'Value': [mape, r2, adjr2]
})
# st.write(metrics_table)
fig.update_layout(
xaxis=dict(title='Date'),
yaxis=dict(title=target_column),
xaxis_tickangle=-30
)
fig.add_annotation(
text=f"MAPE: {mape*100:0.1f}%, Adjr2: {adjr2 *100:.1f}%",
xref="paper",
yref="paper",
x=0.95, # Adjust these values to position the annotation
y=1.2,
showarrow=False,
)
# print("{}{}"*20, len(line_values))
#metrics_table.set_index(['Metric'],inplace=True)
return metrics_table,line_values, fig
def plot_residual_predicted(actual, predicted, df):
df_=df.copy()
df_['Residuals'] = actual - pd.Series(predicted)
df_['StdResidual'] = (df_['Residuals'] - df_['Residuals'].mean()) / df_['Residuals'].std()
# Create a Plotly scatter plot
fig = px.scatter(df_, x=predicted, y='StdResidual', opacity=0.5,color_discrete_sequence=["#11B6BD"])
# Add horizontal lines
fig.add_hline(y=0, line_dash="dash", line_color="darkorange")
fig.add_hline(y=2, line_color="red")
fig.add_hline(y=-2, line_color="red")
fig.update_xaxes(title='Predicted')
fig.update_yaxes(title='Standardized Residuals (Actual - Predicted)')
# Set the same width and height for both figures
fig.update_layout(title='2.3.1 Residuals over Predicted Values', autosize=False, width=600, height=400)
return fig
def residual_distribution(actual, predicted):
Residuals = actual - pd.Series(predicted)
# Create a Seaborn distribution plot
sns.set(style="whitegrid")
plt.figure(figsize=(6, 4))
sns.histplot(Residuals, kde=True, color="#11B6BD")
plt.title('2.3.3 Distribution of Residuals')
plt.xlabel('Residuals')
plt.ylabel('Probability Density')
return plt
def qqplot(actual, predicted):
Residuals = actual - pd.Series(predicted)
Residuals = pd.Series(Residuals)
Resud_std = (Residuals - Residuals.mean()) / Residuals.std()
# Create a QQ plot using Plotly with custom colors
fig = go.Figure()
fig.add_trace(go.Scatter(x=sm.ProbPlot(Resud_std).theoretical_quantiles,
y=sm.ProbPlot(Resud_std).sample_quantiles,
mode='markers',
marker=dict(size=5, color="#11B6BD"),
name='QQ Plot'))
# Add the 45-degree reference line
diagonal_line = go.Scatter(
x=[-2, 2], # Adjust the x values as needed to fit the range of your data
y=[-2, 2], # Adjust the y values accordingly
mode='lines',
line=dict(color='red'), # Customize the line color and style
name=' '
)
fig.add_trace(diagonal_line)
# Customize the layout
fig.update_layout(title='2.3.2 QQ Plot of Residuals',title_x=0.5, autosize=False, width=600, height=400,
xaxis_title='Theoretical Quantiles', yaxis_title='Sample Quantiles')
return fig
|