Snowflake: SQL DENSE_RANK() not working as expected. Would there be an alternative?

I am trying to rank my table based on sku and rolling_inbound but there are quite a few other columns in the table which I believe is why I'm not seeing what I need.

For example:- When I run the code:

SELECT 
    week
    sku,
    units_sold,
    inbound,
    rolling_inbound,
    DENSE_RANK() OVER(PARTITION BY sku, rolling_inbound ORDER BY week) AS sku_rank
FROM table

This is the result I am getting:

But what I am trying to get is the following:

I'm not entirely sure why this is happening, I think it is because the different non-unique rows.

Would there be a solution for this ? or an alternative that will allow me to rank as needed ?

Any help would be much appreciated. Thanks :)

This actually looks to be more of a gaps and islands problem than a ranking. It appears that you want to increment the sku_rank whenever the rolling_inbound changes within a given sku, as ordered by the week.

WITH cte AS (
    SELECT *, CASE WHEN LAG(rolling_inbound)
                  OVER (PARTITION BY sku ORDER BY week) = rolling_inbound
                   THEN 0 ELSE 1 END AS cnt
    FROM yourTable
)

SELECT
    week
    sku,
    units_sold,
    inbound,
    rolling_inbound,
    SUM(cnt) OVER (PARTITION BY sku ORDER BY week) sku_rank
FROM cte
ORDER BY sku, week;