Closed flex567 closed 4 months ago
In one of the latest versions support for enumeration was added, meaning it will try to detect coded columns that only contain values like Yes
, No
or Mild
,Medium
,Severe
etc.
However, I suspect that column 5 in your dataset contains text comments where it's mostly empty, for example with only 2 or 3 non-empty values out of 100 records, then the CSV Lint plug-in will inadvertedly flag it as an enumeration column.
Currently the plug-in counts the unique values of the column, and if it's fewer than 15 (UniqueValuesMax
setting) then it's flagged as enumeration column. However, I think it should also require that at least one or more values must be found more than just 1 single time.
In other words, when it finds for example 7 unique values for a column, but each value is found only once, then it should not be interpreted as a coded value but instead as plain text.
I think the detection of enumeration can be improved, I'll look into this.
I cant replicate the issue and I don't see this error anymore.
After I press Validate: ** error line 6: Column 5 value "Tall car with medium long brown hood. Maybe English, special about it was it had windows like the bottom of Coca-Cola. Nice car, not very cute. Taller then my." is not a valid enumeration member
Not sure what is this error about??