Extraction of data from my timeseries fails with ZeroDivisionError

[vganapathy-lifesize]

I have the following dataset (haven't put the complete one here)
id,date,participants
250,1517344239,2
418,1497884457,6
63,1515513662,3
67,1498667379,2
498,1503235860,2
45,1501160446,10
61,1515016822,3
1,1515169968,2
563,1497884443,8
184,1523390349,3
42,1516111608,3
85,1516095293,2
498,1503531487,2

id - represents virtual meetings (each id identifies a series of meetings that occured. For some could have occurred once every week, some every day some once every n weeks and so on). I am treating each meeting as a series. Hope thats fine
date - time (UTC) when the meeting happened
participants - number of participants that joined the meeting

I am trying to see if I can apply time series analysis to this kind of a data to predict number of users who might join a future occurrence. I am exploring tsfresh to see if I can get more features out for my data which can improve my prediction (its a regression problem). This is the context

series = pd.read_csv('meetings.data')
X = series[[col for col in series.columns if col != 'participants']]
y = series['participants']
extracted_features = extract_features(X, column_id='id', column_sort='date')

This results in following error

Traceback (most recent call last):
  File "tsfresh_predict.py", line 16, in <module>
    extracted_features = extract_features(X, column_id='id', column_sort='date')
  File "/Users/vganapathy/mcufe/dev3/lib/python3.6/site-packages/tsfresh/feature_extraction/extraction.py", line 152, in extract_features
    distributor=distributor)
  File "/Users/vganapathy/mcufe/dev3/lib/python3.6/site-packages/tsfresh/feature_extraction/extraction.py", line 233, in _do_extraction
    function_kwargs=kwargs)
  File "/Users/vganapathy/mcufe/dev3/lib/python3.6/site-packages/tsfresh/utilities/distribution.py", line 142, in map_reduce
    total_number_of_expected_results = math.ceil(data_length / chunk_size)
ZeroDivisionError: division by zero

In my data time of the meeting is the only parameter I have. Just would like to know what is causing this exception.

[nikhase]

You have the ZeroDivisionError because you pass no data. You extract features on the X dataframe, but this solely consists of the columns id and date.
You need to use your series dataframe and specify a column_value like this:

extracted_features = extract_features(series, column_id='id', column_sort='date', column_value='participants')

[MaxBenChrist]

Actually, we should add a test + warning message for this case. Will see that I add a test + error message