Hello,
I have created a file based data set that has a date type column and I am trying to demo a simple Visualization with Year, Month, Qty as bar chart. This works fine. Now I also want to show how easy it was to add Forecast so I right click and click on Add Statistics and click on Forecast and it I get an error on the canvas: cannot assemble with duplicate keys.
While building the Visual, I expanded the Date field which now shows Year, Month, Qtr etc. so I dragged the Year and Month to the canvas and I get a nice Bar chart showing quantity by year/month. So I am unable to understand why the Forecast is giving this error.
Any pointers?
------------------------------------------------------------------------------------
State: HY000. Code: 43119. [nQSError: 43119] Query Failed:
Error Message: cannot assemble with duplicate keys
Traceback (most recent call last):
File "/bi/fasttmp/obis2/tmp/nQS_PAF_44735_1_75781848.TMP", line 94, in <module>
return_dat = obi_execute_script(dat, columnMetadata, args)
File "/bi/fasttmp/obis2/tmp/nQS_PAF_44735_1_75781848.TMP", line 53, in obi_execute_script
output = TimeSeries.timeSeries_columnLevel(dat, target, args)
File "/bi/app/fmw/bi/bifoundation/advanced_analytics/python_scripts/obipy/obitimeseries/TimeSeries.py", line 78, in timeSeries_columnLevel
output = timeutils.partitionedOperation(df, args, target)
File "/bi/app/fmw/bi/bifoundation/advanced_analytics/python_scripts/obipy/obisys/timeutils.py", line 2289, in partitionedOperation
output = single_partition_time_series(dat, actual_free_time_col_names, target_col, target, algorithm,
File "/bi/app/fmw/bi/bifoundation/advanced_analytics/python_scripts/obipy/obisys/timeutils.py", line 2437, in single_partition_time_series
prediction = time_series_function(feature_df, actual_free_time_col_names, target, algorithm, forecast_periods,
File "/bi/app/fmw/bi/bifoundation/advanced_analytics/python_scripts/obipy/obisys/timeutils.py", line 2490, in time_series_function
feature_df, col_names, date_format, timeStamp, col_names_groupby, agg_dict = prepare_time_col_format(feature_df, time_cols, target)
File "/bi/app/fmw/bi/bifoundation/advanced_analytics/python_scripts/obipy/obisys/timeutils.py", line 2703, in prepare_time_col_format
feature_df['timeDate'] = pd.to_datetime(time_data['timeDate'].astype(str),
File "build/bdist.linux-x86_64/egg/pandas/core/tools/datetimes.py", line 890, in to_datetime
File "build/bdist.linux-x86_64/egg/pandas/core/tools/datetimes.py", line 975, in _assemble_from_unit_mappings
ValueError: cannot assemble with duplicate keys (HY000)

------------------------------------------------------------------------------------

Thanks,
Manish