Handle scenario where dataset's last data point is complete and must not be ignored

[sachinkbansal]

Discussed in https://github.com/cuebook/CueObserve/discussions/172

^{Originally posted by PhilippeDo October 8, 2021} I have now this error when I tried to use prophet

{"d2b9be2a-6d27-4e61-a577-3d99892e94e0": {"dimVal": null, "error": "{\"message\": \"single positional indexer is out-of-bounds\", \"stackTrace\": \"Traceback (most recent call last):\\n File \\\"/code/ops/tasks/anomalyDetection.py\\\", line 56, in anomalyService\\n result = detect(df, granularity, detectionRuleType, anomalyDef)\\n File \\\"/code/ops/tasks/anomalyDetection.py\\\", line 29, in detect\\n return prophetDetect(df, granularity)\\n File \\\"/code/ops/tasks/detectionTypes/prophet.py\\\", line 47, in prophetDetect\\n lastISO = df.iloc[-1][\\\"ds\\\"]\\n File \\\"/opt/venv/lib/python3.7/site-packages/pandas/core/indexing.py\\\", line 895, in __getitem__\\n return self._getitem_axis(maybe_callable, axis=axis)\\n File \\\"/opt/venv/lib/python3.7/site-packages/pandas/core/indexing.py\\\", line 1501, in _getitem_axis\\n self._validate_integer(key, axis)\\n File \\\"/opt/venv/lib/python3.7/site-packages/pandas/core/indexing.py\\\", line 1444, in _validate_integer\\n raise IndexError(\\\"single positional indexer is out-of-bounds\\\")\\nIndexError: single positional indexer is out-of-bounds\\n\"}", "success": false}}

I used the following in DATASET section

select DATE_FORMAT(date, '%Y-%m-%d %H') as Date,
pagelt from pageloadtime2

and then when I run, the dataset looks liks this. It is a minimal dataset withou dimension, with only a measure and a timestamp