Excel data imported to KNIME that consists of stacked tables with multiple headers throughout

My data is being imported to knime via an excel reader and the issue I am experiencing is with transforming the imported data into a knime friendly format with one row as headers and data following. The current format has labels and column headers from rows 0-4 then data following. At about row 22 the same labels and column headers are once again listed (for a different client this time) and this takes up another four rows with new data following. The excel sheet is basically stacked tables, each with their own headers. How can I transform this so that the workflow can recognize when one of these 4 row dividers has been reached, take necessary actions in formatting, then repeat the process? The lengths of the data will vary so I cannot hardcode certain row numbers to be removed; however, the columns will always be the same from table to table. Thanks

@cameron767 welcome to the knime forum. You can take a look at this example, you identify the start of your tables and mark them as blocks and then import them in a loop

Hi @cameron767 , if you are using KNIME 5.x, the Table Splitter node could be your friend here too.

To turn one data table into multiple tables, chain multiple Table Splitter nodes with each being provided the conditions (such as the column headings to look for)

1 Like