This page illustrates some errors that you may encounter when using Duco and how to deal with them.
Data format errors
Let's take as an example the following matching process. If you want to try the example used in this page, you can set up the process using the files "payments_demo_data_errors_A.csv" and "payments_demo_data_errors_B1.csv" included in the zip file.
And this is how the fields are aligned:
If you run the process, you'll see results as follows:
Note that the navigation bar on the left of the screen indicates "0 errors".
Now try to run the matching process again using input files "payments_demo_data_errors_A.csv" and "payments_demo_data_errors_B2.csv". The second file contains some format errors.
- Click on Process → Overview
- Click on Run process → Submit new data
- Select file "payments_demo_data_errors_A.csv" for "Data set A"
- Select file "payments_demo_data_errors_B2.csv" for "Data set B"
- Click Start Run
Note that the navigation bar on the left of the screen now indicates "4 errors".
- Click on "Errors" to see more details
You can open the input file "payments_demo_data_errors_B2.csv" with a text editor (for example notepad on Windows) to see its content.
This screen shows the error for the two data sets. There are no errors in "Data set A", but there are some errors in "Data set B". If you click on the ID for an error, you'll see more details about the error. Let's go through the errors one by one.
Error in line 2
There is a problem in line 2, column 1. i.e. the value "15/Jan/2013". If you go back to the Data input screen, you can see that the "Payment Date" field has a "Date" format. More specifically "dd/MM/yyyy". In other words day/month/year all expressed as numbers. For example: "15/01/2013" to indicate "15 January 2013" . The value in line 2 does not use this format.
If you think your input data is going to use both date formats, you can just to add an additional date format for the "Payment date" field in the Data inputs screen.
Error in line 3
There is a problem in line 3, column 5. I.e., the value "USD 70,000.00". If you go back to the Data input screen, you can see that the "Amount" field has a "Decimal" format. The value that is causing the problem contains a mixture of letters and numbers so Duco reports it as an error.
How to deal with this? If the problem is in the input data, you need to fix the data and perhaps have a discussion with who is producing the data.
If, in your case, it's normal that the currency symbol and the amount are both included in the same field value, you'll have to change the process set up to handle this. In this specific example, you could:
- Go to the Data inputs screen and change the format of the "Amount" field to "text"
- Go to the Match fields screen and use a matching rule such as: "Set to the first number embedded in..."
Error in line 4
If you look closely at the line in the input file, you'll notice that the value "GBP" is surrounded with only one quote character. This is just an incorrect CSV format. If the input file contains values that are quoted incorrectly Duco can't determine when a value ends and the following value begins.
Error in line 7
The problem here is that the line contains an extra value. This process is set up to expect five values for each record. Namely: 1) Payment Date, 2) Payer, 3) Receiver, 4) Currency, 5) Amount. This records contains also a sixth value: "Error 1".
How to deal with errors in general
When errors occur, the root cause is usually one of the following:
- Your data input is not set up correctly
- Your input data contains unexpected values
- Your input data is incorrect
Incorrect set-up of Data Inputs
The page Data inputs explains in detail how to set up your data inputs. To set up a data input you need to upload some sample data. Duco can automatically detect the format of the data in your input. To do this accurately set up the process using a representative initial data set. If you find that Duco has detected data formats in an unexpected way, it is possible you set up the process using sample data that was not large enough or representative. If this is the case, it might be easier to create a new process using a representative data set rather than trying to fix the data format for each field.
Otherwise, you can edit your data inputs set up to fix the incorrect data formats. The page Data inputs explains in detail how to set up data formats.
Unexpected values in input data
Another common source of errors is where the input data itself is incorrect, containing unexpected values. For example, you may have a field called "amount" that typically contains values like "1,000,000.00", but may also occasionally contain values like "GBP 1,000,000.00". There are two common solutions for these kinds of problems.
The first solution is to solve the problem within Duco and update your process set-up to handle this data. In this example, you could use the rule "Set to the first number embedded in..." in the Match fields screen. A previous section of this document already describes this.
The second solution is to solve the problem outside of Duco. In this example, there might be a problem with the way in which the input data is being generated. You'll have to speak with the person responsible for generating the data and arrange for them that to send correct and consistent data.
Incorrect input data
Sometimes the input data is simply incorrect. For example, a record may contain too many or too few columns. A problem of this type cannot be solved within Duco. The only solution is to fix the input data at the source.