Hugging Face's logo

Spaces:
rishirajpathak
/
hospital-ai-forecasting
Build error

App Files Community
rishirajpathak commited on
Commit
3ebb301
·
1 Parent(s): a541153

Remove unnecessary files: prediction CSVs and unused scripts

Browse files
Files changed (10) hide show
  1. data/predictions_Current_Weekend.csv +0 -55
  2. data/predictions_Next_24_Hours.csv +0 -25
  3. data/predictions_Next_48_Hours.csv +0 -49
  4. data/predictions_Next_48_Hours_Auto.csv +0 -49
  5. data/predictions_Next_Week_7_Days.csv +0 -169
  6. scripts/compare_predictions.py +0 -192
  7. scripts/data_generator.py +0 -123
  8. scripts/emergency_admissions_predictor.py +0 -147
  9. scripts/predict.py +0 -47
  10. scripts/visualize.py +0 -219
data/predictions_Current_Weekend.csv DELETED
@@ -1,55 +0,0 @@
1
- datetime,predicted_emergency_admissions,predicted_icu_demand,predicted_staff_workload
2
- 2026-01-03 20:01:54.727429,1.8511904761904763,0.6837791,1.0812849
3
- 2026-01-03 21:01:54.727429,1.8511904761904763,0.0015314387,0.57848483
4
- 2026-01-03 22:01:54.727429,1.8511904761904763,-0.0030087454,0.40669277
5
- 2026-01-03 23:01:54.727429,1.8511904761904763,0.03673846,0.009536105
6
- 2026-01-04 00:01:54.727429,1.8511904761904763,0.07041023,0.3309968
7
- 2026-01-04 01:01:54.727429,1.8511904761904763,-0.016362969,-0.10803197
8
- 2026-01-04 02:01:54.727429,1.8511904761904763,0.14771965,0.5044662
9
- 2026-01-04 03:01:54.727429,1.8511904761904763,0.035325162,0.17477345
10
- 2026-01-04 04:01:54.727429,1.8511904761904763,0.03769674,0.025916977
11
- 2026-01-04 05:01:54.727429,1.8511904761904763,0.08433261,0.15435499
12
- 2026-01-04 06:01:54.727429,1.8511904761904763,0.04150059,0.09162672
13
- 2026-01-04 07:01:54.727429,1.8511904761904763,0.01106802,0.26552853
14
- 2026-01-04 08:01:54.727429,1.8511904761904763,0.061104525,0.1673679
15
- 2026-01-04 09:01:54.727429,1.8511904761904763,0.0007876241,0.17770171
16
- 2026-01-04 10:01:54.727429,1.8511904761904763,0.0797624,0.027998284
17
- 2026-01-04 11:01:54.727429,1.8511904761904763,0.008823868,0.2589002
18
- 2026-01-04 12:01:54.727429,1.8511904761904763,-0.019359384,0.22089961
19
- 2026-01-04 13:01:54.727429,1.8511904761904763,-0.046255715,0.3585882
20
- 2026-01-04 14:01:54.727429,1.8511904761904763,0.29936665,1.0840256
21
- 2026-01-04 15:01:54.727429,1.8511904761904763,0.38378263,1.2051085
22
- 2026-01-04 16:01:54.727429,1.8511904761904763,0.25184977,0.6845033
23
- 2026-01-04 17:01:54.727429,1.8511904761904763,0.46332866,0.94264126
24
- 2026-01-04 18:01:54.727429,1.8511904761904763,0.23085001,0.64406663
25
- 2026-01-04 19:01:54.727429,1.8511904761904763,0.22248834,0.8695313
26
- 2026-01-04 20:01:54.727429,1.8511904761904763,-0.0671031,0.48716596
27
- 2026-01-04 21:01:54.727429,1.8511904761904763,0.41112992,0.7704077
28
- 2026-01-04 22:01:54.727429,1.8511904761904763,0.000824114,0.2685413
29
- 2026-01-04 23:01:54.727429,1.8511904761904763,0.0040384126,0.64567906
30
- 2026-01-05 00:01:54.727429,1.8511904761904763,-0.013324468,0.2843667
31
- 2026-01-05 01:01:54.727429,1.8511904761904763,-0.06265203,0.0230423
32
- 2026-01-05 02:01:54.727429,1.8511904761904763,-0.046783715,0.31315714
33
- 2026-01-05 03:01:54.727429,1.8511904761904763,-0.020877209,0.15604892
34
- 2026-01-05 04:01:54.727429,1.8511904761904763,0.013056498,0.24097472
35
- 2026-01-05 05:01:54.727429,1.8511904761904763,0.04414215,0.03931477
36
- 2026-01-05 06:01:54.727429,1.8511904761904763,0.37369907,0.009717901
37
- 2026-01-05 07:01:54.727429,1.8511904761904763,-0.011729559,0.21620716
38
- 2026-01-05 08:01:54.727429,1.8511904761904763,0.0076462706,0.7143712
39
- 2026-01-05 09:01:54.727429,1.8511904761904763,-0.017555328,0.5357925
40
- 2026-01-05 10:01:54.727429,1.8511904761904763,0.110423654,0.5550465
41
- 2026-01-05 11:01:54.727429,1.8511904761904763,-0.034608245,0.094390094
42
- 2026-01-05 12:01:54.727429,1.8511904761904763,0.25977656,0.5042274
43
- 2026-01-05 13:01:54.727429,1.8511904761904763,0.55863285,1.336583
44
- 2026-01-05 14:01:54.727429,1.8511904761904763,0.3547941,0.6865614
45
- 2026-01-05 15:01:54.727429,1.8511904761904763,0.33942354,1.35424
46
- 2026-01-05 16:01:54.727429,1.8511904761904763,0.017039398,0.34210208
47
- 2026-01-05 17:01:54.727429,1.8511904761904763,0.2546735,0.9247641
48
- 2026-01-05 18:01:54.727429,1.8511904761904763,0.41423064,0.6199723
49
- 2026-01-05 19:01:54.727429,1.8511904761904763,0.25366646,0.9620646
50
- 2026-01-05 20:01:54.727429,1.8511904761904763,0.8006692,1.4931285
51
- 2026-01-05 21:01:54.727429,1.8511904761904763,0.90760255,2.1086142
52
- 2026-01-05 22:01:54.727429,1.8511904761904763,0.49078903,1.4082191
53
- 2026-01-05 23:01:54.727429,1.8511904761904763,0.7337919,2.274773
54
- 2026-01-06 00:01:54.727429,1.8511904761904763,0.27126345,0.8296799
55
- 2026-01-06 01:01:54.727429,1.8511904761904763,0.26952648,0.8072494
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
data/predictions_Next_24_Hours.csv DELETED
@@ -1,25 +0,0 @@
1
- datetime,predicted_emergency_admissions,predicted_icu_demand,predicted_staff_workload
2
- 2026-01-03 20:01:54.727429,1.8511904761904763,-0.046783715,0.31315714
3
- 2026-01-03 21:01:54.727429,1.8511904761904763,-0.020877209,0.15604892
4
- 2026-01-03 22:01:54.727429,1.8511904761904763,0.013056498,0.24097472
5
- 2026-01-03 23:01:54.727429,1.8511904761904763,0.04414215,0.03931477
6
- 2026-01-04 00:01:54.727429,1.8511904761904763,0.37369907,0.009717901
7
- 2026-01-04 01:01:54.727429,1.8511904761904763,-0.011729559,0.21620716
8
- 2026-01-04 02:01:54.727429,1.8511904761904763,0.0076462706,0.7143712
9
- 2026-01-04 03:01:54.727429,1.8511904761904763,-0.017555328,0.5357925
10
- 2026-01-04 04:01:54.727429,1.8511904761904763,0.110423654,0.5550465
11
- 2026-01-04 05:01:54.727429,1.8511904761904763,-0.034608245,0.094390094
12
- 2026-01-04 06:01:54.727429,1.8511904761904763,0.25977656,0.5042274
13
- 2026-01-04 07:01:54.727429,1.8511904761904763,0.55863285,1.336583
14
- 2026-01-04 08:01:54.727429,1.8511904761904763,0.3547941,0.6865614
15
- 2026-01-04 09:01:54.727429,1.8511904761904763,0.33942354,1.35424
16
- 2026-01-04 10:01:54.727429,1.8511904761904763,0.017039398,0.34210208
17
- 2026-01-04 11:01:54.727429,1.8511904761904763,0.2546735,0.9247641
18
- 2026-01-04 12:01:54.727429,1.8511904761904763,0.41423064,0.6199723
19
- 2026-01-04 13:01:54.727429,1.8511904761904763,0.25366646,0.9620646
20
- 2026-01-04 14:01:54.727429,1.8511904761904763,0.8006692,1.4931285
21
- 2026-01-04 15:01:54.727429,1.8511904761904763,0.90760255,2.1086142
22
- 2026-01-04 16:01:54.727429,1.8511904761904763,0.49078903,1.4082191
23
- 2026-01-04 17:01:54.727429,1.8511904761904763,0.7337919,2.274773
24
- 2026-01-04 18:01:54.727429,1.8511904761904763,0.27126345,0.8296799
25
- 2026-01-04 19:01:54.727429,1.8511904761904763,0.26952648,0.8072494
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
data/predictions_Next_48_Hours.csv DELETED
@@ -1,49 +0,0 @@
1
- datetime,predicted_emergency_admissions,predicted_icu_demand,predicted_staff_workload
2
- 2026-01-03 20:01:54.727429,1.8511904761904763,0.14771965,0.5044662
3
- 2026-01-03 21:01:54.727429,1.8511904761904763,0.035325162,0.17477345
4
- 2026-01-03 22:01:54.727429,1.8511904761904763,0.03769674,0.025916977
5
- 2026-01-03 23:01:54.727429,1.8511904761904763,0.08433261,0.15435499
6
- 2026-01-04 00:01:54.727429,1.8511904761904763,0.04150059,0.09162672
7
- 2026-01-04 01:01:54.727429,1.8511904761904763,0.01106802,0.26552853
8
- 2026-01-04 02:01:54.727429,1.8511904761904763,0.061104525,0.1673679
9
- 2026-01-04 03:01:54.727429,1.8511904761904763,0.0007876241,0.17770171
10
- 2026-01-04 04:01:54.727429,1.8511904761904763,0.0797624,0.027998284
11
- 2026-01-04 05:01:54.727429,1.8511904761904763,0.008823868,0.2589002
12
- 2026-01-04 06:01:54.727429,1.8511904761904763,-0.019359384,0.22089961
13
- 2026-01-04 07:01:54.727429,1.8511904761904763,-0.046255715,0.3585882
14
- 2026-01-04 08:01:54.727429,1.8511904761904763,0.29936665,1.0840256
15
- 2026-01-04 09:01:54.727429,1.8511904761904763,0.38378263,1.2051085
16
- 2026-01-04 10:01:54.727429,1.8511904761904763,0.25184977,0.6845033
17
- 2026-01-04 11:01:54.727429,1.8511904761904763,0.46332866,0.94264126
18
- 2026-01-04 12:01:54.727429,1.8511904761904763,0.23085001,0.64406663
19
- 2026-01-04 13:01:54.727429,1.8511904761904763,0.22248834,0.8695313
20
- 2026-01-04 14:01:54.727429,1.8511904761904763,-0.0671031,0.48716596
21
- 2026-01-04 15:01:54.727429,1.8511904761904763,0.41112992,0.7704077
22
- 2026-01-04 16:01:54.727429,1.8511904761904763,0.000824114,0.2685413
23
- 2026-01-04 17:01:54.727429,1.8511904761904763,0.0040384126,0.64567906
24
- 2026-01-04 18:01:54.727429,1.8511904761904763,-0.013324468,0.2843667
25
- 2026-01-04 19:01:54.727429,1.8511904761904763,-0.06265203,0.0230423
26
- 2026-01-04 20:01:54.727429,1.8511904761904763,-0.046783715,0.31315714
27
- 2026-01-04 21:01:54.727429,1.8511904761904763,-0.020877209,0.15604892
28
- 2026-01-04 22:01:54.727429,1.8511904761904763,0.013056498,0.24097472
29
- 2026-01-04 23:01:54.727429,1.8511904761904763,0.04414215,0.03931477
30
- 2026-01-05 00:01:54.727429,1.8511904761904763,0.37369907,0.009717901
31
- 2026-01-05 01:01:54.727429,1.8511904761904763,-0.011729559,0.21620716
32
- 2026-01-05 02:01:54.727429,1.8511904761904763,0.0076462706,0.7143712
33
- 2026-01-05 03:01:54.727429,1.8511904761904763,-0.017555328,0.5357925
34
- 2026-01-05 04:01:54.727429,1.8511904761904763,0.110423654,0.5550465
35
- 2026-01-05 05:01:54.727429,1.8511904761904763,-0.034608245,0.094390094
36
- 2026-01-05 06:01:54.727429,1.8511904761904763,0.25977656,0.5042274
37
- 2026-01-05 07:01:54.727429,1.8511904761904763,0.55863285,1.336583
38
- 2026-01-05 08:01:54.727429,1.8511904761904763,0.3547941,0.6865614
39
- 2026-01-05 09:01:54.727429,1.8511904761904763,0.33942354,1.35424
40
- 2026-01-05 10:01:54.727429,1.8511904761904763,0.017039398,0.34210208
41
- 2026-01-05 11:01:54.727429,1.8511904761904763,0.2546735,0.9247641
42
- 2026-01-05 12:01:54.727429,1.8511904761904763,0.41423064,0.6199723
43
- 2026-01-05 13:01:54.727429,1.8511904761904763,0.25366646,0.9620646
44
- 2026-01-05 14:01:54.727429,1.8511904761904763,0.8006692,1.4931285
45
- 2026-01-05 15:01:54.727429,1.8511904761904763,0.90760255,2.1086142
46
- 2026-01-05 16:01:54.727429,1.8511904761904763,0.49078903,1.4082191
47
- 2026-01-05 17:01:54.727429,1.8511904761904763,0.7337919,2.274773
48
- 2026-01-05 18:01:54.727429,1.8511904761904763,0.27126345,0.8296799
49
- 2026-01-05 19:01:54.727429,1.8511904761904763,0.26952648,0.8072494
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
data/predictions_Next_48_Hours_Auto.csv DELETED
@@ -1,49 +0,0 @@
1
- datetime,predicted_emergency_admissions,predicted_icu_demand,predicted_staff_workload
2
- 2026-01-03 20:01:54.727429,1.8511904761904763,0.14771965,0.5044662
3
- 2026-01-03 21:01:54.727429,1.8511904761904763,0.035325162,0.17477345
4
- 2026-01-03 22:01:54.727429,1.8511904761904763,0.03769674,0.025916977
5
- 2026-01-03 23:01:54.727429,1.8511904761904763,0.08433261,0.15435499
6
- 2026-01-04 00:01:54.727429,1.8511904761904763,0.04150059,0.09162672
7
- 2026-01-04 01:01:54.727429,1.8511904761904763,0.01106802,0.26552853
8
- 2026-01-04 02:01:54.727429,1.8511904761904763,0.061104525,0.1673679
9
- 2026-01-04 03:01:54.727429,1.8511904761904763,0.0007876241,0.17770171
10
- 2026-01-04 04:01:54.727429,1.8511904761904763,0.0797624,0.027998284
11
- 2026-01-04 05:01:54.727429,1.8511904761904763,0.008823868,0.2589002
12
- 2026-01-04 06:01:54.727429,1.8511904761904763,-0.019359384,0.22089961
13
- 2026-01-04 07:01:54.727429,1.8511904761904763,-0.046255715,0.3585882
14
- 2026-01-04 08:01:54.727429,1.8511904761904763,0.29936665,1.0840256
15
- 2026-01-04 09:01:54.727429,1.8511904761904763,0.38378263,1.2051085
16
- 2026-01-04 10:01:54.727429,1.8511904761904763,0.25184977,0.6845033
17
- 2026-01-04 11:01:54.727429,1.8511904761904763,0.46332866,0.94264126
18
- 2026-01-04 12:01:54.727429,1.8511904761904763,0.23085001,0.64406663
19
- 2026-01-04 13:01:54.727429,1.8511904761904763,0.22248834,0.8695313
20
- 2026-01-04 14:01:54.727429,1.8511904761904763,-0.0671031,0.48716596
21
- 2026-01-04 15:01:54.727429,1.8511904761904763,0.41112992,0.7704077
22
- 2026-01-04 16:01:54.727429,1.8511904761904763,0.000824114,0.2685413
23
- 2026-01-04 17:01:54.727429,1.8511904761904763,0.0040384126,0.64567906
24
- 2026-01-04 18:01:54.727429,1.8511904761904763,-0.013324468,0.2843667
25
- 2026-01-04 19:01:54.727429,1.8511904761904763,-0.06265203,0.0230423
26
- 2026-01-04 20:01:54.727429,1.8511904761904763,-0.046783715,0.31315714
27
- 2026-01-04 21:01:54.727429,1.8511904761904763,-0.020877209,0.15604892
28
- 2026-01-04 22:01:54.727429,1.8511904761904763,0.013056498,0.24097472
29
- 2026-01-04 23:01:54.727429,1.8511904761904763,0.04414215,0.03931477
30
- 2026-01-05 00:01:54.727429,1.8511904761904763,0.37369907,0.009717901
31
- 2026-01-05 01:01:54.727429,1.8511904761904763,-0.011729559,0.21620716
32
- 2026-01-05 02:01:54.727429,1.8511904761904763,0.0076462706,0.7143712
33
- 2026-01-05 03:01:54.727429,1.8511904761904763,-0.017555328,0.5357925
34
- 2026-01-05 04:01:54.727429,1.8511904761904763,0.110423654,0.5550465
35
- 2026-01-05 05:01:54.727429,1.8511904761904763,-0.034608245,0.094390094
36
- 2026-01-05 06:01:54.727429,1.8511904761904763,0.25977656,0.5042274
37
- 2026-01-05 07:01:54.727429,1.8511904761904763,0.55863285,1.336583
38
- 2026-01-05 08:01:54.727429,1.8511904761904763,0.3547941,0.6865614
39
- 2026-01-05 09:01:54.727429,1.8511904761904763,0.33942354,1.35424
40
- 2026-01-05 10:01:54.727429,1.8511904761904763,0.017039398,0.34210208
41
- 2026-01-05 11:01:54.727429,1.8511904761904763,0.2546735,0.9247641
42
- 2026-01-05 12:01:54.727429,1.8511904761904763,0.41423064,0.6199723
43
- 2026-01-05 13:01:54.727429,1.8511904761904763,0.25366646,0.9620646
44
- 2026-01-05 14:01:54.727429,1.8511904761904763,0.8006692,1.4931285
45
- 2026-01-05 15:01:54.727429,1.8511904761904763,0.90760255,2.1086142
46
- 2026-01-05 16:01:54.727429,1.8511904761904763,0.49078903,1.4082191
47
- 2026-01-05 17:01:54.727429,1.8511904761904763,0.7337919,2.274773
48
- 2026-01-05 18:01:54.727429,1.8511904761904763,0.27126345,0.8296799
49
- 2026-01-05 19:01:54.727429,1.8511904761904763,0.26952648,0.8072494
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
data/predictions_Next_Week_7_Days.csv DELETED
@@ -1,169 +0,0 @@
1
- datetime,predicted_emergency_admissions,predicted_icu_demand,predicted_staff_workload
2
- 2026-01-03 20:01:54.727429,1.8511904761904763,0.16711123,-0.05676522
3
- 2026-01-03 21:01:54.727429,1.8511904761904763,0.23809563,0.5613027
4
- 2026-01-03 22:01:54.727429,1.8511904761904763,-0.015030537,0.3560763
5
- 2026-01-03 23:01:54.727429,1.8511904761904763,-0.027441263,-0.06664164
6
- 2026-01-04 00:01:54.727429,1.8511904761904763,0.24516271,0.84016794
7
- 2026-01-04 01:01:54.727429,1.8511904761904763,0.026039017,0.020360386
8
- 2026-01-04 02:01:54.727429,1.8511904761904763,0.41497785,0.8753252
9
- 2026-01-04 03:01:54.727429,1.8511904761904763,-0.037282396,0.048277006
10
- 2026-01-04 04:01:54.727429,1.8511904761904763,0.5218017,0.6505697
11
- 2026-01-04 05:01:54.727429,1.8511904761904763,-0.0010243563,-0.026307909
12
- 2026-01-04 06:01:54.727429,1.8511904761904763,0.73867816,1.4324422
13
- 2026-01-04 07:01:54.727429,1.8511904761904763,0.10135795,0.36957538
14
- 2026-01-04 08:01:54.727429,1.8511904761904763,0.8841107,1.7499824
15
- 2026-01-04 09:01:54.727429,1.8511904761904763,0.41575915,1.0143486
16
- 2026-01-04 10:01:54.727429,1.8511904761904763,0.01585423,0.5576176
17
- 2026-01-04 11:01:54.727429,1.8511904761904763,0.680214,1.5180483
18
- 2026-01-04 12:01:54.727429,1.8511904761904763,0.98137516,2.2043133
19
- 2026-01-04 13:01:54.727429,1.8511904761904763,0.9434725,1.514512
20
- 2026-01-04 14:01:54.727429,1.8511904761904763,0.64557,1.1800373
21
- 2026-01-04 15:01:54.727429,1.8511904761904763,0.5114943,0.9769659
22
- 2026-01-04 16:01:54.727429,1.8511904761904763,0.110038064,0.19120356
23
- 2026-01-04 17:01:54.727429,1.8511904761904763,0.044258464,0.5329291
24
- 2026-01-04 18:01:54.727429,1.8511904761904763,-0.04493141,0.17622958
25
- 2026-01-04 19:01:54.727429,1.8511904761904763,0.15261811,0.43444055
26
- 2026-01-04 20:01:54.727429,1.8511904761904763,-0.04499425,0.679378
27
- 2026-01-04 21:01:54.727429,1.8511904761904763,0.060131337,-0.008271349
28
- 2026-01-04 22:01:54.727429,1.8511904761904763,0.23572803,0.85977805
29
- 2026-01-04 23:01:54.727429,1.8511904761904763,-0.05951807,0.2288866
30
- 2026-01-05 00:01:54.727429,1.8511904761904763,0.0013675694,0.12297985
31
- 2026-01-05 01:01:54.727429,1.8511904761904763,0.04160405,0.6536843
32
- 2026-01-05 02:01:54.727429,1.8511904761904763,0.059862226,0.27190435
33
- 2026-01-05 03:01:54.727429,1.8511904761904763,0.020274146,0.04551112
34
- 2026-01-05 04:01:54.727429,1.8511904761904763,0.029762378,0.24318805
35
- 2026-01-05 05:01:54.727429,1.8511904761904763,-0.015885219,0.32022694
36
- 2026-01-05 06:01:54.727429,1.8511904761904763,0.039947193,0.46455717
37
- 2026-01-05 07:01:54.727429,1.8511904761904763,0.53780574,0.8961083
38
- 2026-01-05 08:01:54.727429,1.8511904761904763,0.4426041,0.9832277
39
- 2026-01-05 09:01:54.727429,1.8511904761904763,0.08772796,0.4400076
40
- 2026-01-05 10:01:54.727429,1.8511904761904763,0.40821946,0.4261972
41
- 2026-01-05 11:01:54.727429,1.8511904761904763,-0.0017448067,0.5513622
42
- 2026-01-05 12:01:54.727429,1.8511904761904763,-0.023329588,0.11865444
43
- 2026-01-05 13:01:54.727429,1.8511904761904763,0.017126635,0.5288313
44
- 2026-01-05 14:01:54.727429,1.8511904761904763,0.0042125885,0.19921544
45
- 2026-01-05 15:01:54.727429,1.8511904761904763,0.30775723,0.69166034
46
- 2026-01-05 16:01:54.727429,1.8511904761904763,0.033853892,0.0964785
47
- 2026-01-05 17:01:54.727429,1.8511904761904763,0.30003536,0.70292974
48
- 2026-01-05 18:01:54.727429,1.8511904761904763,0.13987379,0.55798686
49
- 2026-01-05 19:01:54.727429,1.8511904761904763,-0.043660484,0.0058952435
50
- 2026-01-05 20:01:54.727429,1.8511904761904763,0.27068883,0.8263014
51
- 2026-01-05 21:01:54.727429,1.8511904761904763,0.30949056,0.76651716
52
- 2026-01-05 22:01:54.727429,1.8511904761904763,0.121107094,0.45636407
53
- 2026-01-05 23:01:54.727429,1.8511904761904763,0.39968586,0.6624997
54
- 2026-01-06 00:01:54.727429,1.8511904761904763,0.18393953,0.7907505
55
- 2026-01-06 01:01:54.727429,1.8511904761904763,-0.015124797,0.67423564
56
- 2026-01-06 02:01:54.727429,1.8511904761904763,0.0909034,0.18552387
57
- 2026-01-06 03:01:54.727429,1.8511904761904763,0.16159329,0.22145866
58
- 2026-01-06 04:01:54.727429,1.8511904761904763,0.042053267,0.088539615
59
- 2026-01-06 05:01:54.727429,1.8511904761904763,0.31881332,0.6419583
60
- 2026-01-06 06:01:54.727429,1.8511904761904763,-0.00014513712,-0.052177463
61
- 2026-01-06 07:01:54.727429,1.8511904761904763,-0.026864538,0.54857504
62
- 2026-01-06 08:01:54.727429,1.8511904761904763,0.13692711,0.621012
63
- 2026-01-06 09:01:54.727429,1.8511904761904763,0.08245306,0.28434563
64
- 2026-01-06 10:01:54.727429,1.8511904761904763,0.03193872,0.5016791
65
- 2026-01-06 11:01:54.727429,1.8511904761904763,0.0052655507,0.60221523
66
- 2026-01-06 12:01:54.727429,1.8511904761904763,0.6703716,1.6241696
67
- 2026-01-06 13:01:54.727429,1.8511904761904763,0.14163043,0.6446874
68
- 2026-01-06 14:01:54.727429,1.8511904761904763,0.8214776,1.5739459
69
- 2026-01-06 15:01:54.727429,1.8511904761904763,0.30830967,0.687604
70
- 2026-01-06 16:01:54.727429,1.8511904761904763,0.6952881,1.651728
71
- 2026-01-06 17:01:54.727429,1.8511904761904763,0.6772458,1.4237005
72
- 2026-01-06 18:01:54.727429,1.8511904761904763,0.6735665,2.2567267
73
- 2026-01-06 19:01:54.727429,1.8511904761904763,0.6245936,1.2617451
74
- 2026-01-06 20:01:54.727429,1.8511904761904763,0.29235876,0.56069165
75
- 2026-01-06 21:01:54.727429,1.8511904761904763,0.037912857,0.029774772
76
- 2026-01-06 22:01:54.727429,1.8511904761904763,0.0006211057,-0.0024309228
77
- 2026-01-06 23:01:54.727429,1.8511904761904763,0.0125262095,0.49224055
78
- 2026-01-07 00:01:54.727429,1.8511904761904763,0.07738231,0.25968695
79
- 2026-01-07 01:01:54.727429,1.8511904761904763,0.1508888,0.52532744
80
- 2026-01-07 02:01:54.727429,1.8511904761904763,-0.009379032,0.05913456
81
- 2026-01-07 03:01:54.727429,1.8511904761904763,0.028096797,0.13971503
82
- 2026-01-07 04:01:54.727429,1.8511904761904763,-0.0028897722,0.035668496
83
- 2026-01-07 05:01:54.727429,1.8511904761904763,0.005176758,0.019116817
84
- 2026-01-07 06:01:54.727429,1.8511904761904763,0.66443366,0.8893655
85
- 2026-01-07 07:01:54.727429,1.8511904761904763,0.13710034,0.8388018
86
- 2026-01-07 08:01:54.727429,1.8511904761904763,0.66277885,1.2761595
87
- 2026-01-07 09:01:54.727429,1.8511904761904763,0.48067543,1.027115
88
- 2026-01-07 10:01:54.727429,1.8511904761904763,0.66759956,1.2679433
89
- 2026-01-07 11:01:54.727429,1.8511904761904763,1.413016,2.666932
90
- 2026-01-07 12:01:54.727429,1.8511904761904763,0.63570267,1.938356
91
- 2026-01-07 13:01:54.727429,1.8511904761904763,2.0741026,3.3654048
92
- 2026-01-07 14:01:54.727429,1.8511904761904763,0.7550563,0.6475589
93
- 2026-01-07 15:01:54.727429,1.8511904761904763,0.98027694,1.1981444
94
- 2026-01-07 16:01:54.727429,1.8511904761904763,-0.049829707,0.049419228
95
- 2026-01-07 17:01:54.727429,1.8511904761904763,0.37397683,0.7713111
96
- 2026-01-07 18:01:54.727429,1.8511904761904763,0.22953668,0.30292514
97
- 2026-01-07 19:01:54.727429,1.8511904761904763,0.0922635,0.18470708
98
- 2026-01-07 20:01:54.727429,1.8511904761904763,0.23658472,0.48473203
99
- 2026-01-07 21:01:54.727429,1.8511904761904763,0.0059197224,-0.029040812
100
- 2026-01-07 22:01:54.727429,1.8511904761904763,0.18034875,0.47250703
101
- 2026-01-07 23:01:54.727429,1.8511904761904763,0.056129947,0.20020348
102
- 2026-01-08 00:01:54.727429,1.8511904761904763,-0.006729746,0.4568517
103
- 2026-01-08 01:01:54.727429,1.8511904761904763,-0.006285989,0.18559098
104
- 2026-01-08 02:01:54.727429,1.8511904761904763,0.18481287,0.2607244
105
- 2026-01-08 03:01:54.727429,1.8511904761904763,0.046784516,0.0818868
106
- 2026-01-08 04:01:54.727429,1.8511904761904763,0.0637247,-0.040680557
107
- 2026-01-08 05:01:54.727429,1.8511904761904763,0.12827277,0.11450266
108
- 2026-01-08 06:01:54.727429,1.8511904761904763,-0.004476241,0.029179627
109
- 2026-01-08 07:01:54.727429,1.8511904761904763,0.69076735,1.0934484
110
- 2026-01-08 08:01:54.727429,1.8511904761904763,0.20836064,0.38317826
111
- 2026-01-08 09:01:54.727429,1.8511904761904763,0.47516993,1.5735811
112
- 2026-01-08 10:01:54.727429,1.8511904761904763,0.055289235,0.90215987
113
- 2026-01-08 11:01:54.727429,1.8511904761904763,0.14111638,0.62369597
114
- 2026-01-08 12:01:54.727429,1.8511904761904763,0.29123738,0.47810504
115
- 2026-01-08 13:01:54.727429,1.8511904761904763,0.4996144,0.84940016
116
- 2026-01-08 14:01:54.727429,1.8511904761904763,0.6837791,1.0812849
117
- 2026-01-08 15:01:54.727429,1.8511904761904763,0.0015314387,0.57848483
118
- 2026-01-08 16:01:54.727429,1.8511904761904763,-0.0030087454,0.40669277
119
- 2026-01-08 17:01:54.727429,1.8511904761904763,0.03673846,0.009536105
120
- 2026-01-08 18:01:54.727429,1.8511904761904763,0.07041023,0.3309968
121
- 2026-01-08 19:01:54.727429,1.8511904761904763,-0.016362969,-0.10803197
122
- 2026-01-08 20:01:54.727429,1.8511904761904763,0.14771965,0.5044662
123
- 2026-01-08 21:01:54.727429,1.8511904761904763,0.035325162,0.17477345
124
- 2026-01-08 22:01:54.727429,1.8511904761904763,0.03769674,0.025916977
125
- 2026-01-08 23:01:54.727429,1.8511904761904763,0.08433261,0.15435499
126
- 2026-01-09 00:01:54.727429,1.8511904761904763,0.04150059,0.09162672
127
- 2026-01-09 01:01:54.727429,1.8511904761904763,0.01106802,0.26552853
128
- 2026-01-09 02:01:54.727429,1.8511904761904763,0.061104525,0.1673679
129
- 2026-01-09 03:01:54.727429,1.8511904761904763,0.0007876241,0.17770171
130
- 2026-01-09 04:01:54.727429,1.8511904761904763,0.0797624,0.027998284
131
- 2026-01-09 05:01:54.727429,1.8511904761904763,0.008823868,0.2589002
132
- 2026-01-09 06:01:54.727429,1.8511904761904763,-0.019359384,0.22089961
133
- 2026-01-09 07:01:54.727429,1.8511904761904763,-0.046255715,0.3585882
134
- 2026-01-09 08:01:54.727429,1.8511904761904763,0.29936665,1.0840256
135
- 2026-01-09 09:01:54.727429,1.8511904761904763,0.38378263,1.2051085
136
- 2026-01-09 10:01:54.727429,1.8511904761904763,0.25184977,0.6845033
137
- 2026-01-09 11:01:54.727429,1.8511904761904763,0.46332866,0.94264126
138
- 2026-01-09 12:01:54.727429,1.8511904761904763,0.23085001,0.64406663
139
- 2026-01-09 13:01:54.727429,1.8511904761904763,0.22248834,0.8695313
140
- 2026-01-09 14:01:54.727429,1.8511904761904763,-0.0671031,0.48716596
141
- 2026-01-09 15:01:54.727429,1.8511904761904763,0.41112992,0.7704077
142
- 2026-01-09 16:01:54.727429,1.8511904761904763,0.000824114,0.2685413
143
- 2026-01-09 17:01:54.727429,1.8511904761904763,0.0040384126,0.64567906
144
- 2026-01-09 18:01:54.727429,1.8511904761904763,-0.013324468,0.2843667
145
- 2026-01-09 19:01:54.727429,1.8511904761904763,-0.06265203,0.0230423
146
- 2026-01-09 20:01:54.727429,1.8511904761904763,-0.046783715,0.31315714
147
- 2026-01-09 21:01:54.727429,1.8511904761904763,-0.020877209,0.15604892
148
- 2026-01-09 22:01:54.727429,1.8511904761904763,0.013056498,0.24097472
149
- 2026-01-09 23:01:54.727429,1.8511904761904763,0.04414215,0.03931477
150
- 2026-01-10 00:01:54.727429,1.8511904761904763,0.37369907,0.009717901
151
- 2026-01-10 01:01:54.727429,1.8511904761904763,-0.011729559,0.21620716
152
- 2026-01-10 02:01:54.727429,1.8511904761904763,0.0076462706,0.7143712
153
- 2026-01-10 03:01:54.727429,1.8511904761904763,-0.017555328,0.5357925
154
- 2026-01-10 04:01:54.727429,1.8511904761904763,0.110423654,0.5550465
155
- 2026-01-10 05:01:54.727429,1.8511904761904763,-0.034608245,0.094390094
156
- 2026-01-10 06:01:54.727429,1.8511904761904763,0.25977656,0.5042274
157
- 2026-01-10 07:01:54.727429,1.8511904761904763,0.55863285,1.336583
158
- 2026-01-10 08:01:54.727429,1.8511904761904763,0.3547941,0.6865614
159
- 2026-01-10 09:01:54.727429,1.8511904761904763,0.33942354,1.35424
160
- 2026-01-10 10:01:54.727429,1.8511904761904763,0.017039398,0.34210208
161
- 2026-01-10 11:01:54.727429,1.8511904761904763,0.2546735,0.9247641
162
- 2026-01-10 12:01:54.727429,1.8511904761904763,0.41423064,0.6199723
163
- 2026-01-10 13:01:54.727429,1.8511904761904763,0.25366646,0.9620646
164
- 2026-01-10 14:01:54.727429,1.8511904761904763,0.8006692,1.4931285
165
- 2026-01-10 15:01:54.727429,1.8511904761904763,0.90760255,2.1086142
166
- 2026-01-10 16:01:54.727429,1.8511904761904763,0.49078903,1.4082191
167
- 2026-01-10 17:01:54.727429,1.8511904761904763,0.7337919,2.274773
168
- 2026-01-10 18:01:54.727429,1.8511904761904763,0.27126345,0.8296799
169
- 2026-01-10 19:01:54.727429,1.8511904761904763,0.26952648,0.8072494
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
scripts/compare_predictions.py DELETED
@@ -1,192 +0,0 @@
1
- """
2
- Compare predictions across different time periods
3
- """
4
-
5
- import pandas as pd
6
- import matplotlib.pyplot as plt
7
- import seaborn as sns
8
- from glob import glob
9
- import os
10
-
11
- def compare_all_periods():
12
- """Create comparison visualization for all prediction periods"""
13
-
14
- print("Loading all prediction files...")
15
-
16
- # Find all prediction files
17
- pred_files = glob('predictions_*.csv')
18
-
19
- if not pred_files:
20
- print("No prediction files found. Run predictions first!")
21
- return
22
-
23
- fig, axes = plt.subplots(2, 2, figsize=(16, 10))
24
- fig.suptitle('Hospital Prediction Comparison: All Time Periods',
25
- fontsize=16, fontweight='bold', y=0.995)
26
-
27
- colors = sns.color_palette("husl", len(pred_files))
28
-
29
- # 1. Emergency Admissions Comparison
30
- ax1 = axes[0, 0]
31
- for i, file in enumerate(sorted(pred_files)):
32
- df = pd.read_csv(file)
33
- df['datetime'] = pd.to_datetime(df['datetime'])
34
- period_name = file.replace('predictions_', '').replace('.csv', '').replace('_', ' ')
35
-
36
- # Plot first 48 hours for comparison
37
- plot_df = df.head(48)
38
- ax1.plot(range(len(plot_df)), plot_df['predicted_emergency_admissions'],
39
- label=period_name, linewidth=2, alpha=0.8, color=colors[i])
40
-
41
- ax1.set_title('Emergency Admissions Forecast', fontsize=12, fontweight='bold')
42
- ax1.set_xlabel('Hours Ahead')
43
- ax1.set_ylabel('Admissions per Hour')
44
- ax1.legend(loc='best', fontsize=8)
45
- ax1.grid(True, alpha=0.3)
46
-
47
- # 2. ICU Demand Comparison
48
- ax2 = axes[0, 1]
49
- for i, file in enumerate(sorted(pred_files)):
50
- df = pd.read_csv(file)
51
- period_name = file.replace('predictions_', '').replace('.csv', '').replace('_', ' ')
52
- plot_df = df.head(48)
53
- ax2.plot(range(len(plot_df)), plot_df['predicted_icu_demand'],
54
- label=period_name, linewidth=2, alpha=0.8, color=colors[i])
55
-
56
- ax2.axhline(y=20, color='red', linestyle='--', label='ICU Capacity', linewidth=2)
57
- ax2.set_title('ICU Demand Forecast', fontsize=12, fontweight='bold')
58
- ax2.set_xlabel('Hours Ahead')
59
- ax2.set_ylabel('ICU Beds Needed')
60
- ax2.legend(loc='best', fontsize=8)
61
- ax2.grid(True, alpha=0.3)
62
-
63
- # 3. Staff Workload Comparison
64
- ax3 = axes[1, 0]
65
- for i, file in enumerate(sorted(pred_files)):
66
- df = pd.read_csv(file)
67
- period_name = file.replace('predictions_', '').replace('.csv', '').replace('_', ' ')
68
- plot_df = df.head(48)
69
- ax3.plot(range(len(plot_df)), plot_df['predicted_staff_workload'],
70
- label=period_name, linewidth=2, alpha=0.8, color=colors[i])
71
-
72
- ax3.set_title('Staff Workload Forecast', fontsize=12, fontweight='bold')
73
- ax3.set_xlabel('Hours Ahead')
74
- ax3.set_ylabel('Staff Required')
75
- ax3.legend(loc='best', fontsize=8)
76
- ax3.grid(True, alpha=0.3)
77
-
78
- # 4. Summary Statistics Table
79
- ax4 = axes[1, 1]
80
- ax4.axis('off')
81
-
82
- # Create summary table
83
- summary_data = []
84
- for file in sorted(pred_files):
85
- df = pd.read_csv(file)
86
- period_name = file.replace('predictions_', '').replace('.csv', '').replace('_', ' ')
87
-
88
- summary_data.append([
89
- period_name,
90
- len(df),
91
- f"{df['predicted_emergency_admissions'].sum():.0f}",
92
- f"{df['predicted_icu_demand'].max():.1f}",
93
- f"{df['predicted_staff_workload'].max():.0f}"
94
- ])
95
-
96
- table = ax4.table(cellText=summary_data,
97
- colLabels=['Period', 'Hours', 'Total Admits', 'Peak ICU', 'Peak Staff'],
98
- cellLoc='center',
99
- loc='center',
100
- colWidths=[0.25, 0.15, 0.15, 0.15, 0.15])
101
-
102
- table.auto_set_font_size(False)
103
- table.set_fontsize(9)
104
- table.scale(1, 2)
105
-
106
- # Style header
107
- for i in range(5):
108
- table[(0, i)].set_facecolor('#2E86AB')
109
- table[(0, i)].set_text_props(weight='bold', color='white')
110
-
111
- # Style rows
112
- for i in range(1, len(summary_data) + 1):
113
- for j in range(5):
114
- if i % 2 == 0:
115
- table[(i, j)].set_facecolor('#F0F0F0')
116
-
117
- ax4.set_title('Summary Statistics', fontsize=12, fontweight='bold', pad=20)
118
-
119
- plt.tight_layout()
120
- plt.savefig('prediction_comparison.png', dpi=300, bbox_inches='tight')
121
- print("✓ Comparison chart saved as prediction_comparison.png")
122
-
123
- # Create individual period summaries
124
- create_period_summary_cards()
125
-
126
- def create_period_summary_cards():
127
- """Create summary cards for each period"""
128
-
129
- fig, axes = plt.subplots(2, 3, figsize=(15, 8))
130
- fig.suptitle('Prediction Period Summary Cards', fontsize=16, fontweight='bold')
131
- axes = axes.flatten()
132
-
133
- pred_files = sorted(glob('predictions_*.csv'))[:6] # Max 6 cards
134
-
135
- for idx, file in enumerate(pred_files):
136
- if idx >= 6:
137
- break
138
-
139
- ax = axes[idx]
140
- df = pd.read_csv(file)
141
- period_name = file.replace('predictions_', '').replace('.csv', '').replace('_', ' ')
142
-
143
- # Calculate statistics
144
- total_admits = int(df['predicted_emergency_admissions'].sum())
145
- peak_icu = df['predicted_icu_demand'].max()
146
- peak_staff = int(df['predicted_staff_workload'].max())
147
- hours = len(df)
148
-
149
- # Determine status color
150
- status_color = '#06A77D' if peak_icu < 17 else '#F18F01' if peak_icu < 19 else '#E63946'
151
-
152
- # Clear axis
153
- ax.axis('off')
154
- ax.set_xlim(0, 1)
155
- ax.set_ylim(0, 1)
156
-
157
- # Add colored background
158
- ax.add_patch(plt.Rectangle((0.05, 0.05), 0.9, 0.9,
159
- facecolor=status_color, alpha=0.1, edgecolor=status_color, linewidth=2))
160
-
161
- # Add text
162
- ax.text(0.5, 0.85, period_name, ha='center', va='top',
163
- fontsize=12, fontweight='bold', color=status_color)
164
- ax.text(0.5, 0.68, f'{hours} hours', ha='center', va='top', fontsize=10, color='gray')
165
-
166
- ax.text(0.5, 0.52, f'{total_admits}', ha='center', va='center',
167
- fontsize=24, fontweight='bold', color='#2E86AB')
168
- ax.text(0.5, 0.42, 'Total Admissions', ha='center', va='center', fontsize=8, color='gray')
169
-
170
- ax.text(0.25, 0.25, f'{peak_icu:.1f}', ha='center', va='center',
171
- fontsize=18, fontweight='bold', color='#A23B72')
172
- ax.text(0.25, 0.15, 'Peak ICU', ha='center', va='center', fontsize=7, color='gray')
173
-
174
- ax.text(0.75, 0.25, f'{peak_staff}', ha='center', va='center',
175
- fontsize=18, fontweight='bold', color='#F18F01')
176
- ax.text(0.75, 0.15, 'Peak Staff', ha='center', va='center', fontsize=7, color='gray')
177
-
178
- # Hide unused subplots
179
- for idx in range(len(pred_files), 6):
180
- axes[idx].axis('off')
181
-
182
- plt.tight_layout()
183
- plt.savefig('period_summary_cards.png', dpi=300, bbox_inches='tight')
184
- print("✓ Summary cards saved as period_summary_cards.png")
185
-
186
- if __name__ == "__main__":
187
- print("=== Creating Prediction Comparison Visualizations ===\n")
188
- compare_all_periods()
189
- print("\n✓ All comparison visualizations created!")
190
- print("\nGenerated files:")
191
- print(" • prediction_comparison.png - Side-by-side comparison")
192
- print(" • period_summary_cards.png - Quick reference cards")
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
scripts/data_generator.py DELETED
@@ -1,123 +0,0 @@
1
- """
2
- Generate realistic mock hospital data with seasonality and external factors
3
- """
4
-
5
- import pandas as pd
6
- import numpy as np
7
- from datetime import datetime, timedelta
8
- import config
9
-
10
- np.random.seed(42)
11
-
12
- def generate_hospital_data(days=180):
13
- """Generate synthetic hospital data with realistic patterns"""
14
-
15
- # Date range
16
- end_date = datetime.now()
17
- start_date = end_date - timedelta(days=days)
18
- dates = pd.date_range(start=start_date, end=end_date, freq='H')
19
-
20
- df = pd.DataFrame({'datetime': dates})
21
- df['date'] = df['datetime'].dt.date
22
- df['hour'] = df['datetime'].dt.hour
23
- df['day_of_week'] = df['datetime'].dt.dayofweek
24
- df['month'] = df['datetime'].dt.month
25
- df['is_weekend'] = df['day_of_week'].isin([5, 6]).astype(int)
26
-
27
- # Time-based patterns
28
- # More admissions in winter (flu season) and evening hours
29
- seasonal_factor = 1 + 0.3 * np.sin(2 * np.pi * df['month'] / 12 - np.pi/2) # Peak in Dec-Jan
30
- hourly_factor = 1 + 0.4 * np.sin(2 * np.pi * (df['hour'] - 6) / 24) # Peak 6 PM
31
- weekend_factor = 1 + 0.2 * df['is_weekend']
32
-
33
- # External factors (covariates)
34
- df['temperature'] = 20 + 10 * np.sin(2 * np.pi * df['month'] / 12) + np.random.normal(0, 3, len(df))
35
- df['flu_season_index'] = (df['month'].isin([11, 12, 1, 2])).astype(float) * 0.7 + np.random.uniform(0, 0.3, len(df))
36
- df['air_quality_index'] = 50 + 30 * np.random.random(len(df)) + 20 * (df['month'].isin([6, 7, 8])).astype(int)
37
-
38
- # Emergency Admissions (main target)
39
- base_admissions = config.BASE_EMERGENCY_ADMISSIONS / 24 # Per hour
40
- df['emergency_admissions'] = (
41
- base_admissions *
42
- seasonal_factor *
43
- hourly_factor *
44
- weekend_factor *
45
- (1 + 0.2 * df['flu_season_index']) *
46
- np.random.gamma(2, 0.5, len(df))
47
- ).round().astype(int)
48
-
49
- # ICU Demand (15-20% of emergency admissions need ICU)
50
- icu_rate = 0.15 + 0.05 * df['flu_season_index']
51
- df['icu_demand'] = (df['emergency_admissions'] * icu_rate).round().astype(int)
52
- df['icu_demand'] = df['icu_demand'].clip(upper=config.ICU_CAPACITY)
53
-
54
- # Staff Workload (complex: depends on patients + severity)
55
- severity_factor = 1 + 0.3 * df['flu_season_index'] + 0.2 * (df['hour'] >= 18).astype(int)
56
- df['staff_workload'] = (
57
- df['emergency_admissions'] * severity_factor *
58
- (config.STAFF_PER_10_PATIENTS / 10)
59
- ).round().astype(int)
60
-
61
- # Current bed occupancy
62
- df['bed_occupancy'] = np.random.randint(30, 90, len(df))
63
-
64
- # Add some random spikes (simulate outbreaks/accidents)
65
- spike_indices = np.random.choice(len(df), size=int(len(df) * 0.02), replace=False)
66
- df.loc[spike_indices, 'emergency_admissions'] *= np.random.uniform(1.5, 2.5, len(spike_indices))
67
- df.loc[spike_indices, 'icu_demand'] *= np.random.uniform(1.3, 2.0, len(spike_indices))
68
-
69
- # Round values
70
- df['emergency_admissions'] = df['emergency_admissions'].round().astype(int)
71
- df['icu_demand'] = df['icu_demand'].round().astype(int)
72
- df['staff_workload'] = df['staff_workload'].round().astype(int)
73
-
74
- return df
75
-
76
- def add_lag_features(df, target_col, lags):
77
- """Add lag features for time series"""
78
- for lag in lags:
79
- df[f'{target_col}_lag_{lag}h'] = df[target_col].shift(lag)
80
- return df
81
-
82
- def add_rolling_features(df, target_col, windows):
83
- """Add rolling average features"""
84
- for window in windows:
85
- df[f'{target_col}_rolling_{window}h'] = df[target_col].rolling(window=window).mean()
86
- return df
87
-
88
- def prepare_ml_features(df):
89
- """Prepare features for ML models"""
90
-
91
- # Add lag features for key targets
92
- df = add_lag_features(df, 'emergency_admissions', config.LAG_FEATURES)
93
- df = add_lag_features(df, 'icu_demand', config.LAG_FEATURES)
94
-
95
- # Add rolling features
96
- df = add_rolling_features(df, 'emergency_admissions', config.ROLLING_WINDOWS)
97
-
98
- # Drop NaN rows created by lag features
99
- df = df.dropna()
100
-
101
- return df
102
-
103
- if __name__ == "__main__":
104
- # Generate and save data
105
- print("Generating hospital data...")
106
- df = generate_hospital_data(config.DAYS_OF_HISTORICAL_DATA)
107
-
108
- print(f"Generated {len(df)} hourly records over {config.DAYS_OF_HISTORICAL_DATA} days")
109
- print(f"\nData shape: {df.shape}")
110
- print(f"\nSample data:")
111
- print(df.head())
112
-
113
- print(f"\nStatistics:")
114
- print(df[['emergency_admissions', 'icu_demand', 'staff_workload']].describe())
115
-
116
- # Save raw data
117
- df.to_csv('hospital_data.csv', index=False)
118
- print(f"\nData saved to hospital_data.csv")
119
-
120
- # Prepare ML features
121
- df_ml = prepare_ml_features(df.copy())
122
- df_ml.to_csv('hospital_data_ml.csv', index=False)
123
- print(f"ML-ready data saved to hospital_data_ml.csv")
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
scripts/emergency_admissions_predictor.py DELETED
@@ -1,147 +0,0 @@
1
- """
2
- Emergency Admissions Predictor using Amazon Chronos
3
- State-of-the-art pretrained time-series model
4
- """
5
-
6
- import pandas as pd
7
- import numpy as np
8
- import torch
9
- import config
10
- from chronos import ChronosPipeline
11
- import warnings
12
- warnings.filterwarnings('ignore')
13
-
14
- class EmergencyAdmissionsPredictor:
15
- def __init__(self, model_name=config.CHRONOS_MODEL):
16
- """Initialize Chronos model for emergency admissions prediction"""
17
- self.model_name = model_name
18
- self.pipeline = None
19
-
20
- def load_model(self):
21
- """Load pretrained Chronos model"""
22
- print(f"Loading Chronos model: {self.model_name}...")
23
-
24
- # Check if CUDA is available
25
- device = "cuda" if torch.cuda.is_available() else "cpu"
26
- dtype = torch.bfloat16 if device == "cuda" else torch.float32
27
-
28
- self.pipeline = ChronosPipeline.from_pretrained(
29
- self.model_name,
30
- device_map=device,
31
- torch_dtype=dtype,
32
- )
33
- print(f"Model loaded on {device}")
34
-
35
- def predict(self, historical_data, prediction_length=48, num_samples=20):
36
- """
37
- Predict future emergency admissions
38
-
39
- Args:
40
- historical_data: Array or list of historical admissions (hourly)
41
- prediction_length: Hours to predict ahead
42
- num_samples: Number of forecast samples for uncertainty
43
-
44
- Returns:
45
- Dictionary with predictions and confidence intervals
46
- """
47
- if self.pipeline is None:
48
- self.load_model()
49
-
50
- # Convert to tensor
51
- context = torch.tensor(historical_data, dtype=torch.float32)
52
-
53
- # Generate forecast
54
- forecast = self.pipeline.predict(
55
- context=context,
56
- prediction_length=prediction_length,
57
- num_samples=num_samples,
58
- )
59
-
60
- # Calculate statistics
61
- forecast_np = forecast.squeeze().numpy()
62
-
63
- predictions = {
64
- 'mean': forecast_np.mean(axis=0),
65
- 'median': np.median(forecast_np, axis=0),
66
- 'lower_bound': np.percentile(forecast_np, 10, axis=0),
67
- 'upper_bound': np.percentile(forecast_np, 90, axis=0),
68
- 'std': forecast_np.std(axis=0)
69
- }
70
-
71
- return predictions
72
-
73
- def predict_from_dataframe(self, df, context_length=168, prediction_length=48):
74
- """
75
- Predict from a pandas DataFrame
76
-
77
- Args:
78
- df: DataFrame with 'emergency_admissions' column
79
- context_length: Hours of context to use (default: 1 week)
80
- prediction_length: Hours to predict ahead
81
- """
82
- # Get last context_length hours
83
- historical_data = df['emergency_admissions'].tail(context_length).values
84
-
85
- # Predict
86
- predictions = self.predict(historical_data, prediction_length)
87
-
88
- # Create future dates
89
- last_date = df['datetime'].iloc[-1]
90
- future_dates = pd.date_range(
91
- start=last_date + pd.Timedelta(hours=1),
92
- periods=prediction_length,
93
- freq='H'
94
- )
95
-
96
- # Create prediction DataFrame
97
- pred_df = pd.DataFrame({
98
- 'datetime': future_dates,
99
- 'predicted_admissions': predictions['mean'],
100
- 'predicted_admissions_lower': predictions['lower_bound'],
101
- 'predicted_admissions_upper': predictions['upper_bound'],
102
- })
103
-
104
- return pred_df, predictions
105
-
106
- def demo():
107
- """Demo the emergency admissions predictor"""
108
- print("=== Emergency Admissions Predictor Demo ===\n")
109
-
110
- # Load data
111
- print("Loading hospital data...")
112
- df = pd.read_csv('hospital_data.csv')
113
- df['datetime'] = pd.to_datetime(df['datetime'])
114
-
115
- print(f"Data loaded: {len(df)} hours of data")
116
- print(f"Latest admission count: {df['emergency_admissions'].iloc[-1]}")
117
- print(f"Average admissions (last 7 days): {df['emergency_admissions'].tail(168).mean():.2f}\n")
118
-
119
- # Initialize predictor
120
- predictor = EmergencyAdmissionsPredictor()
121
-
122
- # Make prediction
123
- print(f"Predicting next {config.PREDICTION_HORIZON} hours...")
124
- pred_df, predictions = predictor.predict_from_dataframe(
125
- df,
126
- context_length=168, # Use 1 week of context
127
- prediction_length=config.PREDICTION_HORIZON
128
- )
129
-
130
- # Show results
131
- print("\nPredictions for next 24 hours:")
132
- print(pred_df.head(24))
133
-
134
- # Summary statistics
135
- print(f"\nSummary for next 48 hours:")
136
- print(f"Expected total admissions: {predictions['mean'].sum():.0f}")
137
- print(f"Peak hour prediction: {predictions['mean'].max():.0f} admissions")
138
- print(f"Low hour prediction: {predictions['mean'].min():.0f} admissions")
139
-
140
- # Save predictions
141
- pred_df.to_csv('predictions_emergency_admissions.csv', index=False)
142
- print("\nPredictions saved to predictions_emergency_admissions.csv")
143
-
144
- return pred_df, predictions
145
-
146
- if __name__ == "__main__":
147
- demo()
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
scripts/predict.py DELETED
@@ -1,47 +0,0 @@
1
- """
2
- Simple Command-Line Interface for Quick Predictions
3
- """
4
-
5
- import sys
6
- from predict_flexible import FlexiblePredictor
7
-
8
- def main():
9
- if len(sys.argv) < 2:
10
- print("\nUsage: python predict.py <period>")
11
- print("\nAvailable periods:")
12
- print(" 24h - Next 24 hours")
13
- print(" 48h - Next 48 hours (default)")
14
- print(" weekend - Next weekend")
15
- print(" week - Next week (7 days)")
16
- print(" <number> - Custom hours (e.g., 72 for 3 days)")
17
- print("\nExamples:")
18
- print(" python predict.py 24h")
19
- print(" python predict.py weekend")
20
- print(" python predict.py 120 (5 days)")
21
- return
22
-
23
- period = sys.argv[1].lower()
24
-
25
- predictor = FlexiblePredictor()
26
-
27
- if period == '24h' or period == '24':
28
- predictor.predict_next_24_hours()
29
- elif period == '48h' or period == '48':
30
- predictor.predict_next_48_hours()
31
- elif period == 'weekend':
32
- predictor.predict_next_weekend()
33
- elif period == 'week' or period == '7d':
34
- predictor.predict_next_week()
35
- else:
36
- try:
37
- hours = int(period)
38
- if hours > 0 and hours <= 720:
39
- predictor.predict_custom(hours, f"Next {hours} Hours")
40
- else:
41
- print("Error: Hours must be between 1 and 720")
42
- except ValueError:
43
- print(f"Error: Unknown period '{period}'")
44
- print("Use: 24h, 48h, weekend, week, or a number")
45
-
46
- if __name__ == "__main__":
47
- main()
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
scripts/visualize.py DELETED
@@ -1,219 +0,0 @@
1
- """
2
- Visualization Dashboard for Hospital Prediction System
3
- Creates impressive charts for hackathon presentation
4
- """
5
-
6
- import pandas as pd
7
- import matplotlib.pyplot as plt
8
- import seaborn as sns
9
- import numpy as np
10
- from datetime import datetime
11
- import warnings
12
- warnings.filterwarnings('ignore')
13
-
14
- sns.set_style("whitegrid")
15
- plt.rcParams['figure.figsize'] = (15, 10)
16
-
17
- def create_dashboard():
18
- """Create comprehensive visualization dashboard"""
19
-
20
- # Load data
21
- print("Loading data...")
22
- historical = pd.read_csv('hospital_data.csv')
23
- predictions = pd.read_csv('predictions_final.csv')
24
-
25
- historical['datetime'] = pd.to_datetime(historical['datetime'])
26
- predictions['datetime'] = pd.to_datetime(predictions['datetime'])
27
-
28
- # Create figure with subplots
29
- fig = plt.figure(figsize=(16, 12))
30
-
31
- # 1. Emergency Admissions - Historical + Predictions
32
- ax1 = plt.subplot(3, 2, 1)
33
- last_week = historical.tail(168) # Last 7 days
34
- ax1.plot(last_week['datetime'], last_week['emergency_admissions'],
35
- label='Historical', color='#2E86AB', linewidth=2)
36
- ax1.plot(predictions['datetime'], predictions['predicted_emergency_admissions'],
37
- label='Predicted', color='#F24236', linewidth=2, linestyle='--')
38
- ax1.set_title('Emergency Admissions: Historical vs Predicted', fontsize=14, fontweight='bold')
39
- ax1.set_xlabel('Time')
40
- ax1.set_ylabel('Admissions per Hour')
41
- ax1.legend()
42
- ax1.grid(True, alpha=0.3)
43
-
44
- # 2. ICU Demand Forecast
45
- ax2 = plt.subplot(3, 2, 2)
46
- ax2.plot(predictions['datetime'], predictions['predicted_icu_demand'],
47
- color='#A23B72', linewidth=2.5, marker='o', markersize=3)
48
- ax2.axhline(y=20, color='red', linestyle='--', label='ICU Capacity', linewidth=2)
49
- ax2.fill_between(predictions['datetime'], 0, predictions['predicted_icu_demand'],
50
- alpha=0.3, color='#A23B72')
51
- ax2.set_title('ICU Demand Forecast (48 Hours)', fontsize=14, fontweight='bold')
52
- ax2.set_xlabel('Time')
53
- ax2.set_ylabel('ICU Beds Needed')
54
- ax2.legend()
55
- ax2.grid(True, alpha=0.3)
56
-
57
- # 3. Staff Workload Prediction
58
- ax3 = plt.subplot(3, 2, 3)
59
- ax3.plot(predictions['datetime'], predictions['predicted_staff_workload'],
60
- color='#F18F01', linewidth=2.5, marker='s', markersize=3)
61
- ax3.fill_between(predictions['datetime'], 0, predictions['predicted_staff_workload'],
62
- alpha=0.3, color='#F18F01')
63
- ax3.set_title('Staff Workload Forecast', fontsize=14, fontweight='bold')
64
- ax3.set_xlabel('Time')
65
- ax3.set_ylabel('Staff Required')
66
- ax3.grid(True, alpha=0.3)
67
-
68
- # 4. Seasonal Pattern Analysis
69
- ax4 = plt.subplot(3, 2, 4)
70
- monthly_avg = historical.groupby('month')['emergency_admissions'].mean()
71
- months_available = monthly_avg.index.tolist()
72
- month_names = ['Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec']
73
- labels = [month_names[m-1] for m in months_available]
74
- bars = ax4.bar(months_available, monthly_avg, color=sns.color_palette("coolwarm", len(months_available)))
75
- ax4.set_title('Seasonal Admission Pattern', fontsize=14, fontweight='bold')
76
- ax4.set_xlabel('Month')
77
- ax4.set_ylabel('Avg Hourly Admissions')
78
- ax4.set_xticks(months_available)
79
- ax4.set_xticklabels(labels)
80
- ax4.grid(True, alpha=0.3, axis='y')
81
-
82
- # 5. Hourly Pattern (Heatmap style)
83
- ax5 = plt.subplot(3, 2, 5)
84
- hourly_avg = historical.groupby('hour')['emergency_admissions'].mean()
85
- ax5.bar(range(24), hourly_avg, color='#06A77D', alpha=0.7)
86
- ax5.set_title('Daily Pattern: Peak Hours', fontsize=14, fontweight='bold')
87
- ax5.set_xlabel('Hour of Day')
88
- ax5.set_ylabel('Avg Admissions')
89
- ax5.set_xticks(range(0, 24, 3))
90
- ax5.grid(True, alpha=0.3, axis='y')
91
-
92
- # 6. External Factors Impact
93
- ax6 = plt.subplot(3, 2, 6)
94
- last_days = historical.tail(168)
95
- ax6_twin = ax6.twinx()
96
-
97
- ax6.plot(last_days['datetime'], last_days['emergency_admissions'],
98
- label='Admissions', color='#2E86AB', linewidth=2)
99
- ax6_twin.plot(last_days['datetime'], last_days['flu_season_index'],
100
- label='Flu Season Index', color='#E63946', linewidth=2, alpha=0.7)
101
-
102
- ax6.set_title('External Factors: Flu Season Impact', fontsize=14, fontweight='bold')
103
- ax6.set_xlabel('Time')
104
- ax6.set_ylabel('Admissions', color='#2E86AB')
105
- ax6_twin.set_ylabel('Flu Index', color='#E63946')
106
- ax6.legend(loc='upper left')
107
- ax6_twin.legend(loc='upper right')
108
- ax6.grid(True, alpha=0.3)
109
-
110
- plt.tight_layout()
111
- plt.savefig('hospital_dashboard.png', dpi=300, bbox_inches='tight')
112
- print("✓ Dashboard saved as hospital_dashboard.png")
113
-
114
- # Create summary metrics visualization
115
- create_metrics_summary(predictions)
116
-
117
- plt.show()
118
-
119
- def create_metrics_summary(predictions):
120
- """Create a clean metrics summary chart"""
121
-
122
- fig, axes = plt.subplots(2, 2, figsize=(12, 8))
123
- fig.suptitle('Hospital Emergency Metrics - Next 48 Hours',
124
- fontsize=16, fontweight='bold', y=0.98)
125
-
126
- # 1. Total Expected Admissions
127
- ax1 = axes[0, 0]
128
- total_admissions = int(predictions['predicted_emergency_admissions'].sum())
129
- ax1.text(0.5, 0.5, str(total_admissions),
130
- ha='center', va='center', fontsize=60, fontweight='bold', color='#2E86AB')
131
- ax1.text(0.5, 0.2, 'Expected Admissions',
132
- ha='center', va='center', fontsize=14, color='gray')
133
- ax1.axis('off')
134
- ax1.set_facecolor('#F0F0F0')
135
-
136
- # 2. Peak ICU Demand
137
- ax2 = axes[0, 1]
138
- peak_icu = predictions['predicted_icu_demand'].max()
139
- ax2.text(0.5, 0.5, f'{peak_icu:.1f}',
140
- ha='center', va='center', fontsize=60, fontweight='bold', color='#A23B72')
141
- ax2.text(0.5, 0.2, 'Peak ICU Beds',
142
- ha='center', va='center', fontsize=14, color='gray')
143
- ax2.axis('off')
144
- ax2.set_facecolor('#F0F0F0')
145
-
146
- # 3. Status Indicator
147
- ax3 = axes[1, 0]
148
- status = 'NORMAL'
149
- status_color = '#06A77D'
150
- if peak_icu > 17:
151
- status = 'CRITICAL'
152
- status_color = '#E63946'
153
- elif peak_icu > 14:
154
- status = 'ELEVATED'
155
- status_color = '#F18F01'
156
-
157
- ax3.text(0.5, 0.5, status,
158
- ha='center', va='center', fontsize=48, fontweight='bold', color=status_color)
159
- ax3.text(0.5, 0.2, 'Hospital Status',
160
- ha='center', va='center', fontsize=14, color='gray')
161
- ax3.axis('off')
162
- ax3.set_facecolor('#F0F0F0')
163
-
164
- # 4. Peak Staff Required
165
- ax4 = axes[1, 1]
166
- peak_staff = int(np.ceil(predictions['predicted_staff_workload'].max()))
167
- ax4.text(0.5, 0.5, str(peak_staff),
168
- ha='center', va='center', fontsize=60, fontweight='bold', color='#F18F01')
169
- ax4.text(0.5, 0.2, 'Peak Staff Required',
170
- ha='center', va='center', fontsize=14, color='gray')
171
- ax4.axis('off')
172
- ax4.set_facecolor('#F0F0F0')
173
-
174
- plt.tight_layout()
175
- plt.savefig('hospital_metrics.png', dpi=300, bbox_inches='tight')
176
- print("✓ Metrics summary saved as hospital_metrics.png")
177
-
178
- def create_comparison_chart():
179
- """Create model comparison chart"""
180
-
181
- fig, ax = plt.subplots(figsize=(10, 6))
182
-
183
- models = ['Chronos\n(Transformer)', 'XGBoost\nICU', 'XGBoost\nStaff']
184
- accuracy = [92, 67, 58] # Approximate R² scores as percentage
185
- speed = [3, 0.1, 0.1] # Training time in seconds
186
-
187
- x = np.arange(len(models))
188
- width = 0.35
189
-
190
- ax.bar(x - width/2, accuracy, width, label='Accuracy (%)', color='#2E86AB')
191
- ax2 = ax.twinx()
192
- ax2.bar(x + width/2, speed, width, label='Train Time (s)', color='#F18F01')
193
-
194
- ax.set_xlabel('Model', fontsize=12)
195
- ax.set_ylabel('Accuracy (%)', color='#2E86AB', fontsize=12)
196
- ax2.set_ylabel('Training Time (seconds)', color='#F18F01', fontsize=12)
197
- ax.set_title('Model Performance Comparison', fontsize=14, fontweight='bold')
198
- ax.set_xticks(x)
199
- ax.set_xticklabels(models)
200
- ax.legend(loc='upper left')
201
- ax2.legend(loc='upper right')
202
- ax.grid(True, alpha=0.3, axis='y')
203
-
204
- plt.tight_layout()
205
- plt.savefig('model_comparison.png', dpi=300, bbox_inches='tight')
206
- print("✓ Model comparison saved as model_comparison.png")
207
-
208
- if __name__ == "__main__":
209
- print("=== Creating Hospital Prediction Dashboard ===\n")
210
-
211
- create_dashboard()
212
- create_comparison_chart()
213
-
214
- print("\n✓ All visualizations created!")
215
- print("\nGenerated files:")
216
- print(" • hospital_dashboard.png - Main analytics dashboard")
217
- print(" • hospital_metrics.png - Key metrics summary")
218
- print(" • model_comparison.png - Model performance comparison")
219
- print("\nUse these in your hackathon presentation!")