Time-series plots:Full dataset:Truncated dataset:
Basic statistics:- Last two days dropped due to incomple week (2019w46)Full dataset (only dropped first three days):. tabstat merit, s(n mean sd p50 p25 p75 min max) format(%9.1f)
variable | N mean sd p50 p25 p75 min max
-------------+--------------------------------------------------------------------------------
merit | 655.0 729.0 416.5 636.0 533.0 766.0 295.0 4500.0
----------------------------------------------------------------------------------------------
Applied formulas in previous weeks, potential outliers are days have intra-day merits beyond 184 or 1116.
. di 766-533
233
. di 233*1.5
349.5
. di 766+349.5
1115.5
. di 533-349.5
183.5
There are
52 outliers (beyond 1116 or 184) in full dataset, in total.
. count if (merit >= 1116 | merit <= 184) & merit != .
52
Those days are:
. list id merit date if (merit >= 1116 | merit <= 184) & merit != .
+-------------------------+
| id merit date |
|-------------------------|
1. | 3 4500 26jan2018 |
2. | 4 3490 27jan2018 |
3. | 5 3190 28jan2018 |
4. | 6 3800 29jan2018 |
5. | 7 4193 30jan2018 |
|-------------------------|
6. | 8 2821 31jan2018 |
7. | 9 2546 01feb2018 |
8. | 10 2569 02feb2018 |
9. | 11 1868 03feb2018 |
10. | 12 2182 04feb2018 |
|-------------------------|
11. | 13 2078 05feb2018 |
12. | 14 2310 06feb2018 |
13. | 15 2142 07feb2018 |
14. | 16 2143 08feb2018 |
15. | 17 1449 09feb2018 |
|-------------------------|
16. | 18 1748 10feb2018 |
17. | 19 1443 11feb2018 |
18. | 20 1332 12feb2018 |
19. | 21 1580 13feb2018 |
20. | 22 2514 14feb2018 |
|-------------------------|
21. | 23 1992 15feb2018 |
22. | 24 1416 16feb2018 |
23. | 25 1618 17feb2018 |
24. | 26 1293 18feb2018 |
25. | 27 1404 19feb2018 |
|-------------------------|
26. | 28 1170 20feb2018 |
27. | 29 1268 21feb2018 |
28. | 30 1280 22feb2018 |
30. | 32 1410 24feb2018 |
31. | 33 1187 25feb2018 |
|-------------------------|
32. | 34 1392 26feb2018 |
33. | 35 1327 27feb2018 |
35. | 37 1335 01mar2018 |
36. | 38 1706 02mar2018 |
39. | 41 1246 05mar2018 |
|-------------------------|
46. | 48 1355 12mar2018 |
48. | 50 1160 14mar2018 |
49. | 51 1131 15mar2018 |
54. | 56 1324 20mar2018 |
55. | 57 1229 21mar2018 |
|-------------------------|
66. | 68 1258 01apr2018 |
67. | 69 1147 02apr2018 |
151. | 153 1139 25jun2018 |
234. | 236 2464 16sep2018 |
235. | 237 1863 17sep2018 |
|-------------------------|
236. | 238 1295 18sep2018 |
237. | 239 1271 19sep2018 |
349. | 351 1162 09jan2019 |
354. | 356 1128 14jan2019 |
426. | 428 1250 27mar2019 |
|-------------------------|
473. | 475 1151 13may2019 |
502. | 504 1188 11jun2019 |
+-------------------------+
Only
five of them occured in 2019, on 09/1/2019 (1162), 14/1/2019 (1128), 27/3/2019 (1250), 13/5/2019 (1151), and 11/6/2019 (1188).
Note that the latest spike on 13/11/2019 at 1476 merits changed hands is surely a potential outlier but for this week that day dropped becauase it belongs to the incomplete week (2019w46).
Truncated dataset (first 25 days dropped):. tabstat merit, s(n mean sd p50 p25 p75 min max) format(%9.1f)
variable | N mean sd p50 p25 p75 min max
-------------+--------------------------------------------------------------------------------
merit | 631.0 667.6 216.2 627.0 530.0 742.0 295.0 2464.0
----------------------------------------------------------------------------------------------
Applied same formulas I used in earlier analyses, potential outliers are days have intra-day merits beyond
212 or
1060.
. di 742-530
212
. di 212*1.5
318
. di 742+318
1060
. di 530-318
212
There are
36 outliers in total, only
seven of them occured in 2019, on 04/01/2019 (1083) , 09/1/2019 (1162), 14/01/2019 (1128) , 27/3/2019 (1250), 13/5/2019 (1151), 11/6/2019 (1188), and 21/10/2019 (1082).
. count if (merit >= 1060 | merit <= 212) & merit != .
36
List of those 36 outliers in truncated dataset
. list id merit date if (merit >= 1060 | merit <= 212) & merit != .
+-------------------------+
| id merit date |
|-------------------------|
1. | 27 1404 19feb2018 |
2. | 28 1170 20feb2018 |
3. | 29 1268 21feb2018 |
4. | 30 1280 22feb2018 |
6. | 32 1410 24feb2018 |
|-------------------------|
7. | 33 1187 25feb2018 |
8. | 34 1392 26feb2018 |
9. | 35 1327 27feb2018 |
11. | 37 1335 01mar2018 |
12. | 38 1706 02mar2018 |
|-------------------------|
13. | 39 1090 03mar2018 |
15. | 41 1246 05mar2018 |
16. | 42 1075 06mar2018 |
17. | 43 1111 07mar2018 |
21. | 47 1092 11mar2018 |
|-------------------------|
22. | 48 1355 12mar2018 |
24. | 50 1160 14mar2018 |
25. | 51 1131 15mar2018 |
30. | 56 1324 20mar2018 |
31. | 57 1229 21mar2018 |
|-------------------------|
42. | 68 1258 01apr2018 |
43. | 69 1147 02apr2018 |
44. | 70 1081 03apr2018 |
45. | 71 1062 04apr2018 |
127. | 153 1139 25jun2018 |
|-------------------------|
210. | 236 2464 16sep2018 |
211. | 237 1863 17sep2018 |
212. | 238 1295 18sep2018 |
213. | 239 1271 19sep2018 |
320. | 346 1083 04jan2019 |
|-------------------------|
325. | 351 1162 09jan2019 |
330. | 356 1128 14jan2019 |
402. | 428 1250 27mar2019 |
449. | 475 1151 13may2019 |
478. | 504 1188 11jun2019 |
|-------------------------|
610. | 636 1082 21oct2019 |
+-------------------------+