Common Pitfalls

Having trouble analyzing your data? The following pitfalls could cause an error message or insufficient results:

• Replacing blank or missing values with placeholders, such as NULL, false, or n/a. Emcien provides faster, more accurate results when missing values are left blank.
• Loading unbanded (or unbinned) numerical data. Unbanded numerical data often results in few connections.

  Need help binning you data? You can try our automated data prep module, Bandit (BETA). If you need further help please contact the Emcien team at [email protected]

• Loading files that are not in a CSV format, including files using the .csv file extension but that are not comma delimited.
• Transforming data into RDF or Boolean values (1s and 0s) reduces the effectiveness of Emcien's analysis.
• Shifted or improperly-entered data in the receipt format
• Using the incorrect date and time format. Check out the Date and Time section of our Preparing Your Data article for more information.
• Using negative prices in the item_price or transaction_price columns in the receipt format. Emcien does not support negative prices.

Many of these errors can be overcome without any data cleanup or transformation. If your analysis results in a “parse” error and there are only a limited number of lines with errors, the analysis can continue. Simply navigate back to the “New Analysis” page, open the “Override Defaults” button, and select “Allow Limited Invalid Lines”. 

This will remove those invalid lines and allow the analysis to continue.