Repository navigation
Expand file tree
/
Copy pathmain.py
More file actions
508 lines (455 loc) · 28.2 KB
/
Copy pathmain.py
File metadata and controls
508 lines (455 loc) · 28.2 KB
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
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
import time
import pandas as pd
import streamlit as st
from PIL import Image
from deta import Deta
from Shell_EDA import Shell_EDA
from Shell_Ensemble import Shell_Ensemble
from Shell_LightGBM import Shell_LGBM
from Shell_SARIMAX import Shell_SARIMAX
def main():
deta = Deta(st.secrets["data_key"])
deta_img_drive = deta.Drive("Image")
data_drive = deta.Drive("Data")
shell_logo = deta_img_drive.get("shell_logo.png")
shell_logo_img = Image.open(shell_logo)
main_c1, main_c2, main_c3 = st.columns([1, 5, 1])
with main_c2:
image_c1, image_c2, image_c3 = st.columns([1, 3, 1])
title_c1, title_c2, title_c3 = st.columns([1, 5, 1])
sb_c1, sb_c2, sb_c3 = st.columns([1, 3, 1])
with image_c2:
st.image(shell_logo_img, width=200, use_column_width="auto")
with title_c2:
st.title("Royal Dutch Shell")
with sb_c2:
st.subheader("Cash Flow Forecasting")
st.write("Please provide the following information to generate a cash flow forecast:")
uploaded_file = st.file_uploader("Choose a CSV file")
if uploaded_file is not None:
df = pd.read_csv(uploaded_file)
csv_data = df.to_csv(index=False).encode('utf-8')
data_drive.put("eda_dataset.csv", csv_data)
data_drive.put("train_dataset.csv", csv_data)
df.to_csv('static/eda_dataset.csv')
with st.container():
ms = ""
file_output = pd.DataFrame()
ms_c1, ms_c2, ms_c3 = st.columns([1, 2, 1])
with ms_c2:
model_selection = st.selectbox("Model Selection:", ["", "SARIMAX", "LightGBM", "Ensemble"])
if model_selection == "Ensemble":
col1_param, col2_param = st.columns(2)
if st.checkbox("LightGBM Grid Search"):
lgbm_grid_search = True
else:
lgbm_grid_search = False
with col1_param:
st.write("LightGBM Parameters:")
num_leaves = st.text_input("num_leaves:", placeholder="Default Best Parameter: 31")
max_depth = st.text_input("max_depth:", placeholder="Default Best Parameter: 5")
learning_rate = st.text_input("learning_rate:", placeholder="Default Best Parameter: 0.01")
subsample = st.text_input("subsample:", placeholder="Default Best Parameter: 31")
colsample_bytree = st.text_input("colsample_bytree:", placeholder="Default Best Parameter: 0.6")
reg_alpha = st.text_input("reg_alpha:", placeholder="Default Best Parameter: 0.5")
reg_lambda = st.text_input("reg_lambda:", placeholder="Default Best Parameter: 0.1")
n_estimators = st.text_input("n_estimators:", placeholder="Default Best Parameter: 200")
random_state = st.text_input("random_state:", placeholder="Default Best Parameter: 42")
num_iterations = st.text_input("num_iterations:", placeholder="Default Best Parameter: 200")
try:
lightgbm_parameters = {"num_leaves": int(num_leaves),
"max_depth": int(max_depth),
"learning_rate": float(learning_rate),
"subsample": float(subsample),
"colsample_bytree": float(colsample_bytree),
"reg_alpha": float(reg_alpha),
"reg_lambda": float(reg_lambda),
"n_estimators": int(n_estimators),
"random_state": int(random_state),
"num_iterations": int(num_iterations)}
except:
pass
if st.checkbox("SARIMAX Grid Search"):
sarimax_grid_search = True
else:
sarimax_grid_search = False
with col2_param:
st.write("SARIMAX Parameters:")
st.write("Order")
p = st.text_input("p:", placeholder="Default Best Parameter: 1")
q = st.text_input("q:", placeholder="Default Best Parameter: 0")
d = st.text_input("d:", placeholder="Default Best Parameter: 1")
st.write("Seasonal Order")
P = st.text_input("P:", placeholder="Default Best Parameter: 2")
D = st.text_input("D:", placeholder="Default Best Parameter: 2")
Q = st.text_input("Q:", placeholder="Default Best Parameter: 2")
s = st.text_input("s:", placeholder="Default Best Parameter: 4")
try:
sarimax_params = [int(p), int(q), int(d), int(P), int(D), int(Q), int(s)]
except:
pass
elif model_selection == "LightGBM":
if st.checkbox("LightGBM Grid Search"):
lgbm_grid_search = True
else:
lgbm_grid_search = False
st.write("LightGBM Parameters:")
num_leaves = st.text_input("num_leaves:", placeholder="Default Best Parameter: 31")
max_depth = st.text_input("max_depth:", placeholder="Default Best Parameter: 5")
learning_rate = st.text_input("learning_rate:", placeholder="Default Best Parameter: 0.01")
subsample = st.text_input("subsample:", placeholder="Default Best Parameter: 31")
colsample_bytree = st.text_input("colsample_bytree:", placeholder="Default Best Parameter: 0.6")
reg_alpha = st.text_input("reg_alpha:", placeholder="Default Best Parameter: 0.5")
reg_lambda = st.text_input("reg_lambda:", placeholder="Default Best Parameter: 0.1")
n_estimators = st.text_input("n_estimators:", placeholder="Default Best Parameter: 200")
random_state = st.text_input("random_state:", placeholder="Default Best Parameter: 42")
num_iterations = st.text_input("num_iterations:", placeholder="Default Best Parameter: 200")
try:
lightgbm_parameters = {"num_leaves": int(num_leaves),
"max_depth": int(max_depth),
"learning_rate": float(learning_rate),
"subsample": float(subsample),
"colsample_bytree": float(colsample_bytree),
"reg_alpha": float(reg_alpha),
"reg_lambda": float(reg_lambda),
"n_estimators": int(n_estimators),
"random_state": int(random_state),
"num_iterations": int(num_iterations)}
except:
pass
elif model_selection == "SARIMAX":
if st.checkbox("SARIMAX Grid Search"):
sarimax_grid_search = True
else:
sarimax_grid_search = False
st.write("SARIMAX Parameters:")
st.write("Order")
p = st.text_input("p:", placeholder="Default Best Parameter: 1")
q = st.text_input("q:", placeholder="Default Best Parameter: 0")
d = st.text_input("d:", placeholder="Default Best Parameter: 1")
st.write("Seasonal Order")
P = st.text_input("P:", placeholder="Default Best Parameter: 2")
D = st.text_input("D:", placeholder="Default Best Parameter: 2")
Q = st.text_input("Q:", placeholder="Default Best Parameter: 2")
s = st.text_input("s:", placeholder="Default Best Parameter: 4")
try:
sarimax_params = [int(p), int(q), int(d), int(P), int(D), int(Q), int(s)]
except:
pass
st.write("Training Start/End Dates:")
col1_date, col2_date = st.columns(2)
disabled_1 = False
with col1_date:
training_start_date_sarimax = pd.to_datetime(st.date_input("Training Start Date for SARIMAX:",disabled=disabled_1))
if st.checkbox("Disable Training End Date For SARIMAX?"):
disabled_sarimax = True
else:
disabled_sarimax = False
training_end_date_sarimax = pd.to_datetime(
st.date_input("Training End Date for SARIMAX:", disabled=disabled_sarimax)
)
with col2_date:
training_end_date_lightgbm = pd.to_datetime(st.date_input("Training End Date for LightGBM:",disabled=disabled_1))
if st.checkbox("Disable Training End Date For LightGBM?"):
disabled_lgbm = True
else:
disabled_lgbm = False
training_start_date_lightgbm = pd.to_datetime(
st.date_input("Training Start Date for LightGBM:", disabled=disabled_lgbm)
)
st.write("Test Start/End Dates:")
col1_date_test, col2_date_test, col3_date_test = st.columns([1, 3, 1])
with col2_date_test:
st.write("Test Start/End Dates:")
test_start_date = pd.to_datetime(st.date_input("Test Start Date:"))
test_end_date = pd.to_datetime(st.date_input("Test End Date:"))
analysis_flag = st.empty()
b1_test, b2_test, b3_test = st.columns([2, 3, 2])
with b2_test:
if st.button("Explanatory Data Analysis",disabled = disabled_1):
analysis_flag.value = "EDA"
if st.button("Forecast the Future Cashflow!",disabled = disabled_1):
analysis_flag.value = "Forecast"
if analysis_flag.value == "EDA":
disabled_1 = True
# Perform EDA and display results
shell_data_eda = Shell_EDA(filepath=data_drive,
train_set_name="eda_dataset.csv",
column_name="Net Cashflow from Operations",
start_date="2021-01-01",
end_date=None,
output_path=f"static", )
shell_data_eda.read_data()
acf_pacf_path = shell_data_eda.acf_pacf()
seasonal_decomp_text = shell_data_eda.decompose()
info_text = shell_data_eda.information()
mean_std_plot_path = shell_data_eda.mean_std_plot()
df_test_text = shell_data_eda.DF_test()
seasonal_decomp_path_2 = shell_data_eda.seasonal_decomp()
order_1_path, order_2_path, order_3_path = shell_data_eda.ordered_pacf_acf()
st.title("Explanatory Data Analysis")
with st.container():
st.header("ACF/PACF Plot")
st.image(acf_pacf_path)
with st.container():
st.header("Seasonal Decomposition Plot")
st.text_area("Seasonal Decomposition Plot", seasonal_decomp_text, height=330, label_visibility="hidden")
with st.container():
st.header("Information")
st.text_area("Information", info_text, height=400, label_visibility="hidden")
with st.container():
st.header("Mean/Standard Deviation Plot")
st.image(mean_std_plot_path)
with st.container():
st.header("Seasonal Decomposition Plot")
st.image(seasonal_decomp_path_2)
with st.container():
st.header("DF Test")
st.text_area("DF Test", df_test_text, height=250, label_visibility="hidden")
with st.container():
st.header("ACF/PACF Plot with Order 0")
st.image(order_1_path)
st.header("ACF/PACF Plot with Order 1")
st.image(order_2_path)
st.header("ACF/PACF Plot with Order 2")
st.image(order_3_path)
st.cache_data.clear()
elif analysis_flag.value == "Forecast":
if model_selection == "SARIMAX":
ms = "SARIMAX"
elif model_selection == "LightGBM":
ms = "LightGBM"
elif model_selection == "Ensemble":
ms = "Ensemble"
if ms == "Ensemble":
shell_sarimax_class = Shell_SARIMAX(filepath=data_drive,
train_set_name="train_dataset.csv",
column_name="Net Cashflow from Operations",
start_date=training_start_date_sarimax,
end_date=training_end_date_sarimax,
prediction_start_date=test_start_date,
prediction_end_date=test_end_date)
shell_sarimax_class.create_train_set()
shell_sarimax_class.seasonal_decomposition()
shell_sarimax_class.create_train_test_exog_endog()
if sarimax_grid_search:
st.info('SARIMAX Grid Search Started', icon="ℹ️")
time.sleep(0.5)
shell_sarimax_class.SARIMAX_gridsearch()
st.success('SARIMAX Grid Search Finished', icon="✅")
time.sleep(0.5)
else:
st.info('Loading SARIMAX Best Parameters', icon="ℹ️")
time.sleep(0.5)
shell_sarimax_class.best_params = sarimax_params
st.success('Loaded SARIMAX Best Parameters', icon="✅")
time.sleep(0.5)
st.info('SARIMAX Training is Starting', icon="ℹ️")
time.sleep(0.5)
shell_sarimax_class.SARIMAX_train_test()
st.success('SARIMAX is Trained Succesfully!', icon="✅")
time.sleep(0.5)
st.info('SARIMAX Forecast Started', icon="ℹ️")
time.sleep(0.5)
shell_sarimax_class_forecast_output = shell_sarimax_class.Sarimax_Forecast()
st.success('SARIMAX Forecast is Succesfull!', icon="✅")
time.sleep(0.5)
shell_lgbm_class = Shell_LGBM(filepath=data_drive, train_set_name="train_dataset.csv",
column_name="Net Cashflow from Operations",
start_date=training_start_date_lightgbm,
end_date=training_end_date_lightgbm,
prediction_start_date=test_start_date,
prediction_end_date=test_end_date,
param_set={'objective': ['regression'],
'boosting_type': ["goss"],
'num_leaves': [15, 31, 63, 127, 255],
'max_depth': [3, 5, 7, 15, 31],
'learning_rate': [0.1, 0.01, 0.001],
'subsample': [0.8, 0.6, 1.0],
'colsample_bytree': [0.8, 0.6, 1.0],
'reg_alpha': [0.0, 0.1, 0.5],
'reg_lambda': [0.0, 0.1, 0.5],
'n_estimators': [100, 200, 500],
'random_state': [42],
'num_iterations': [100, 200, 500]})
shell_lgbm_class.create_train_set()
shell_lgbm_class.create_train_test_exog_endog()
if lgbm_grid_search:
st.info('LightGBM Grid Search Started', icon="ℹ️")
time.sleep(0.5)
shell_lgbm_class.LGBM_GridSearch()
st.success('LightGBM Grid Search Finished', icon="✅")
time.sleep(0.5)
else:
st.info('Loading LightGBM Best Parameters', icon="ℹ️")
time.sleep(0.5)
param_set = {'objective': 'regression',
'boosting_type': "goss",
'num_leaves': lightgbm_parameters["num_leaves"],
'max_depth': lightgbm_parameters["max_depth"],
'learning_rate': lightgbm_parameters["learning_rate"],
'subsample': lightgbm_parameters["subsample"],
'colsample_bytree': lightgbm_parameters["colsample_bytree"],
'reg_alpha': lightgbm_parameters["reg_alpha"],
'reg_lambda': lightgbm_parameters["reg_lambda"],
'n_estimators': lightgbm_parameters["n_estimators"],
'random_state': lightgbm_parameters["random_state"],
'num_iterations': lightgbm_parameters["num_iterations"]}
print(param_set)
st.success('Loaded LightGBM Best Parameters', icon="✅")
time.sleep(0.5)
shell_lgbm_class.best_params = param_set
st.info('LightGBM Training is Starting', icon="ℹ️")
time.sleep(0.5)
shell_lgbm_class.Train_LightGBM()
st.success('LightGBM is Trained Succesfully!', icon="✅")
time.sleep(0.5)
st.info('LightGBM Forecast Started', icon="ℹ️")
time.sleep(0.5)
shell_lgbm_class_forecast_output = shell_lgbm_class.Forecast_LightGBM()
st.success('LightGBM Forecast is Succesfull!', icon="✅")
time.sleep(0.5)
sarimax_output = pd.DataFrame(columns=["Date", "Net Cashflow from Operations"])
lgbm_output = pd.DataFrame(columns=["Date", "Net Cashflow from Operations"])
sarimax_output["Date"] = shell_sarimax_class_forecast_output.index
sarimax_output["Net Cashflow from Operations"] = shell_sarimax_class_forecast_output.values
lgbm_output["Date"] = shell_lgbm_class_forecast_output.index
lgbm_output["Net Cashflow from Operations"] = shell_lgbm_class_forecast_output.values
shell_sarimax_class_forecast_output_csv = shell_sarimax_class_forecast_output.to_csv(index=False)
data_drive.put("submission_sarimax.csv", shell_sarimax_class_forecast_output_csv)
shell_lgbm_class_forecast_output_csv = shell_lgbm_class_forecast_output.to_csv(index=False)
data_drive.put("submission_lgbm.csv", shell_lgbm_class_forecast_output_csv)
shell_ensemble_class = Shell_Ensemble(submission_SARIMAX=shell_sarimax_class_forecast_output,
submission_LGBM=shell_lgbm_class_forecast_output,
file_path=shell_sarimax_class.filepath,
file_name="submission_ensemble.csv")
ensemble_output = shell_ensemble_class.Ensemble()
elif ms == "SARIMAX":
shell_sarimax_class = Shell_SARIMAX(filepath=data_drive,
train_set_name="train_dataset.csv",
column_name="Net Cashflow from Operations",
start_date=training_start_date_sarimax,
end_date=training_end_date_sarimax,
prediction_start_date=test_start_date,
prediction_end_date=test_end_date)
shell_sarimax_class.create_train_set()
shell_sarimax_class.seasonal_decomposition()
shell_sarimax_class.create_train_test_exog_endog()
if sarimax_grid_search:
st.info('SARIMAX Grid Search Started', icon="ℹ️")
time.sleep(0.5)
shell_sarimax_class.SARIMAX_gridsearch()
st.success('SARIMAX Grid Search Finished', icon="✅")
time.sleep(0.5)
else:
st.info('Loading SARIMAX Best Parameters', icon="ℹ️")
time.sleep(0.5)
shell_sarimax_class.best_params = sarimax_params
st.success('Loaded SARIMAX Best Parameters', icon="✅")
time.sleep(0.5)
st.info('SARIMAX Training is Starting', icon="ℹ️")
time.sleep(0.5)
shell_sarimax_class.SARIMAX_train_test()
st.success('SARIMAX is Trained Succesfully!', icon="✅")
time.sleep(0.5)
st.info('SARIMAX Forecast Started', icon="ℹ️")
time.sleep(0.5)
shell_sarimax_class_forecast_output = shell_sarimax_class.Sarimax_Forecast()
st.success('SARIMAX Forecast is Succesfull!', icon="✅")
time.sleep(0.5)
sarimax_output = pd.DataFrame(columns=["Date", "Net Cashflow from Operations"])
sarimax_output["Date"] = shell_sarimax_class_forecast_output.index
sarimax_output["Net Cashflow from Operations"] = shell_sarimax_class_forecast_output.values
sarimax_output.to_csv(f"{shell_sarimax_class.filepath}/submission_sarimax.csv",
index=False)
elif ms == "LightGBM":
shell_lgbm_class = Shell_LGBM(filepath=data_drive, train_set_name="train_dataset.csv",
column_name="Net Cashflow from Operations",
start_date=training_start_date_lightgbm,
end_date=training_end_date_lightgbm,
prediction_start_date=test_start_date,
prediction_end_date=test_end_date,
param_set={'objective': ['regression'],
'boosting_type': ["goss"],
'num_leaves': [15, 31, 63, 127, 255],
'max_depth': [3, 5, 7, 15, 31],
'learning_rate': [0.1, 0.01, 0.001],
'subsample': [0.8, 0.6, 1.0],
'colsample_bytree': [0.8, 0.6, 1.0],
'reg_alpha': [0.0, 0.1, 0.5],
'reg_lambda': [0.0, 0.1, 0.5],
'n_estimators': [100, 200, 500],
'random_state': [42],
'num_iterations': [100, 200, 500]})
shell_lgbm_class.create_train_set()
shell_lgbm_class.create_train_test_exog_endog()
if lgbm_grid_search:
st.info('LightGBM Grid Search Started', icon="ℹ️")
time.sleep(0.5)
shell_lgbm_class.LGBM_GridSearch()
st.success('LightGBM Grid Search Finished', icon="✅")
time.sleep(0.5)
else:
st.info('Loading LightGBM Best Parameters', icon="ℹ️")
time.sleep(0.5)
param_set = {'objective': 'regression',
'boosting_type': "goss",
'num_leaves': lightgbm_parameters["num_leaves"],
'max_depth': lightgbm_parameters["max_depth"],
'learning_rate': lightgbm_parameters["learning_rate"],
'subsample': lightgbm_parameters["subsample"],
'colsample_bytree': lightgbm_parameters["colsample_bytree"],
'reg_alpha': lightgbm_parameters["reg_alpha"],
'reg_lambda': lightgbm_parameters["reg_lambda"],
'n_estimators': lightgbm_parameters["n_estimators"],
'random_state': lightgbm_parameters["random_state"],
'num_iterations': lightgbm_parameters["num_iterations"]}
st.success('Loaded LightGBM Best Parameters', icon="✅")
time.sleep(0.5)
shell_lgbm_class.best_params = param_set
st.info('LightGBM Training is Starting', icon="ℹ️")
time.sleep(0.5)
shell_lgbm_class.Train_LightGBM()
st.success('LightGBM is Trained Succesfully!', icon="✅")
time.sleep(0.5)
st.info('LightGBM Forecast Started', icon="ℹ️")
time.sleep(0.5)
shell_lgbm_class_forecast_output = shell_lgbm_class.Forecast_LightGBM()
st.success('LightGBM Forecast is Succesfull!', icon="✅")
time.sleep(0.5)
lgbm_output = pd.DataFrame(columns=["Date", "Net Cashflow from Operations"])
lgbm_output["Date"] = shell_lgbm_class_forecast_output.index
lgbm_output["Net Cashflow from Operations"] = shell_lgbm_class_forecast_output.values
lgbm_output.to_csv(f"{shell_lgbm_class.filepath}/submission_lgbm.csv", index=False)
if ms == "Ensemble":
file_output = ensemble_output
elif ms == "SARIMAX":
file_output = sarimax_output
elif ms == "LightGBM":
file_output = lgbm_output
@st.cache_data
def convert_df(df):
# IMPORTANT: Cache the conversion to prevent computation on every rerun
return df.to_csv(index=False).encode('utf-8')
def download_button_clicked():
st.success('Downloaded Succesfully!', icon="✅")
time.sleep(0.5)
st.balloons()
files_to_delete = ["eda_dataset.csv", "train_dataset.csv",
"submission_sarimax.csv", "submission_lgbm.csv",
"submission_ensemble.csv"]
result = data_drive.delete_many(files_to_delete);
print("deleted:", result.get("deleted"))
print("failed:", result.get("failed"))
st.cache_data.clear()
forecast_csv = convert_df(file_output)
with b2_test:
st.download_button(
label="Download the Forecast as CSV",
data=forecast_csv,
file_name='forecast.csv',
mime='csv',
on_click=download_button_clicked,
)
if __name__ == "__main__":
main()