Most of the time, you would want to avoid as best you can in handling data with a dynamic structure, simply because it is very rare that it is absolutely necessary to deal with data with varying number of columns.

For example, suppose your raw data contains sales data for each month of the year to date:

However, as the data is a rolling history of the months so far in the year (assuming we are currently at the start of August), there are no columns for months after July yet.

And so, in cases like this, every month's data will have a dynamic number of columns.

Despite that, you can simply transpose this dataset into a structure with a static number of columns. It is always good practice to keep your data with a static number of columns for consistency and reliability.

Dealing with unavoidably dynamic number of columns

However, if because of the unique context of your data or in rare cases where dynamic number of columns are needed, this article will show how you can deal with such cases.

Using the same example as before, normally if were to specify a data range in SolveXia, we would simply enter the following configuration:

However, this won't work next month, when there is a new column for the month of August (as it will be in column I).

And so, in order to indicate a 'dynamic' data range, we first need to make use of the "Create named range" instruction.

And so, we need to configure a named range like so:

And now, you can use an instruction that refers to the named range "Data_Range" instead of manually entering the cell range:

The process will now deal behave as originally intended:

Furthermore, next month (when we have a new column for August's data), the process will still behave as intended, without having to explicitly add in column I into the data range in SolveXia: