Refresh a CSV dataset

My dataset consists of a CSV file that I loaded on S3. But this CSV file is not yet final, and as soon as I add a new column to the CSV file, I cannot refresh the dataset. I then have to start all over again with creating a new dataset, but especially with creating a new analysis.

How can I solve this? This bothers me immensely. I read somewhere that I should first import the CSV into RedShift and then it would work. Is this the only solution?


JOHN DOE;2022-01-01;100
JOE SIXPACK;2022-01-01;500

My json manifest file

    "fileLocations": [
            "URIs": [
    "globalUploadSettings": {
        "format": "CSV",
        "delimiter": ";",
        "containsHeader": "true"

After creating the analysis with all the necessary graphs and data, I notice that I need some additional info. I create a new CSV with an additional column.

JOHN DOE;2022-01-01;100;20
JOE SIXPACK;2022-01-01;500;30

But by adding this extra column, I have to start all over again with a new dataset. But most importantly, I have to recreate the analysis again. This is sometimes a work of several days.

Is there really no better solution?


Hi Ivan,

I had posted a solution which you can test when the schema in your file is changing : Dataset from an S3 folder, cannot add new columns on latest CSV file - #7 by Koushik_Muthanna