CODEWITHSUNDEEP
pythonanacondatime-seriesfeature-extractionargument-matching
Shivani Gupta
Shivani Gupta

Reputation: 1

Unable to extract features from time series data using 'extract_features' method from tsfresh package

I am running the code in Spyder(3.3.3) from Anaconda3 2019.03(Python 3.7.3 64-bit). And using tsfresh 0.11.1

The code I'm running deals with a huge set of time-series data that has sensor data(data of 17 sensors in 17 different files.

(dataset url : https://archive.ics.uci.edu/ml/datasets/Condition+monitoring+of+hydraulic+systems)

The problem is that the execution is getting stuck at the feature extraction step. It is using extract_features method from the tsfresh package to extract features from the data. The 'column_id' and 'column_sort' parameters are of 'str' datatype by default. So, it is showing the following error at that step:

TypeError: unsupported operand type(s) for /: 'str' and 'int'

And when tried to execute only one file out of 17 files, it is showing a different error:

TypeError:Cannot cast array data from dtype('float64') to dtype('U32') according to the rule 'safe'

The code url: https://github.com/zhou100/SensorDefaults/blob/master/Detecting%20and%20Compensating%20Sensor%20Faults.ipynb

Thanks

Upvotes: 0

Views: 556

Answers (1)

nirmalnk
nirmalnk

Reputation: 33

I tried removing time column which actually contains '25/12/2020 14:32:23'.

After that I tried , it worked Reason for error TypeError: unsupported operand type(s) for /: 'str' and 'int',

Upvotes: 0

Related Questions