Data360 Analyze

Issue: I can't figure out what the alternative is for the Delimiter File from Lavastorm. 

Use case:

1. Dataset loaded with Directory List

2. Then transform node is used to split a directory name

3. Then transform node is used to made an if else statement: if (in1.mydata == 'example'):
out1 += in1

4. Delimiter file from (lavastorm) reads data.

Question: The current Delimiter/CSV file input node does not work if I fill in: 'Filename' in field: Input.Data (from Filename). What is necessary to get the same result as with the previous 'Delimiter node'?

In the case where the Filename value is contained in an input field you need to configure the CSV/Delimited node so that the source type is set to Data (from Filename Field)  and the value of the property contains the name of the input field. 

The Data Source is configured using the drop down list (indicated by the red rectangle in the above screenshot). You can select the field from using the context menu's Input Fields option or you can type in the field name. The apostrophes are only required if the field name contains space characters or is a reserved keyword. The field name value is case sensitive.

In my example I leveraged the Split node to select the required filenames instead of using a Transform node. The node can be configured manually or you could apply a filter on the data from the Directory List node in the Data Viewer (I used the 'contains' operator) and then add the node to the canvas - so all the configuration is code-free.

Hi Adrian,

We have tested and we have found the following issue in version Data3Sixty 3.4.2.5049.

We have done the test with two different files. With my first file the 'CSV/Delimited Input' node does not work. With the second test file it does work. So I assume there is a character within the first file which I cannot ignore. When we get the data with the 'CSV/Delimited Input' node we receive the following errors:

In this CSV/Delimited File I have filled the following fields:

What I see in 'Lavastorm' for the same file is that we have used the following RecordDelimiter. 

Then the Lavastorm.DelimiterFile does return the right results. 

The function to add a value to the 'RecordDelimiter' seems to be missing in Data3Sixty? How can I create a work around or work with the node in this case? 

Also good to to mention is that with the current Lavastorm.Delimited.File you see that there are results but when I click on the upper output the data does not load. 

This Delimiter function has to be used many times in our data analysis process. 

Thank you. 

The CSV/Delimited node uses an AutoDetect mechanism to determine the formatting of records rather than a property value.

In this case the node appears to be indicating the problem is that a record contains an unescaped quote character. 

When there is a problem reading the file the data is not available at the output. Have you looked at using one of the options in the FileCharacterSet property, e.g. UTF-8 ?

In this situation it may be better for you to create a support ticket using the 'Ask a question" link at the top of this page as you would be able to upload a sanitized version of the sample.csv file to us.