Percentile, Customers buying more expensive or cheaper items

[ElliotP]

Evening,

I'm trying to work out if customer's purchased weighted by volume more expensive or cheaper dishes.

I have transaction data for a small venue which has four dishes for example with each dish having a different price point. I would like to weight the percentile figure by volume if that makes any sense as so I can work out, out of all of the dishes purchased in the past period, did customer's buy the more expensive or the cheaper dishes.

My thought is to find the k = 0.51 percentile of dish price points by volume but I'm not sure how to bridge that gap. I'm not sure if I'm making sense in my approach, but I feel I want to find the k=0.51 price point paid vs the average price point for example (Dish1Price+Dish2Price+Dish3Price+Dish4Price)/NumberofDishes.

I'm not sure how to create the measure to find the k=0.51 price point paid by volume.

Here is a link to my pbix if that's of any help to anyone:  https://1drv.ms/u/s!At8Q-ZbRnAj8hjIV_LCfjAmx4QJF

[v-shex-msft · Accepted solution]

Hi @ElliotP,

>> I feel I want to find the k=0.51 price point paid vs the average price point for example (Dish1Price+Dish2Price+Dish3Price+Dish4Price)/NumberofDishes.

If you have the measure to calculate the percent, you can drag it to visual level filter. filter type to "advanced filter", show item when value field to "is", fill in 0.51 to textbox and apply the filte.

If above is not help, please provide more detailed informaiton to explain your requirement.

Regards,

Xiaoxin Sheng

[ElliotP]

Thanks for the response.

I'm trying to work out the average price point paid weighted by the volume of orders. I think i'll just weight each price point by the proportion of the total volume and do it that way.

[ElliotP]

Thoughts?