4 Replies Latest reply on Jul 21, 2017 8:36 AM by Isaac Briseno

# How to analyze spreadsheet on chunks/clusters of 10 and 20 records

Hello All,

I'm new to Tableau, I'm using Desktop version 10 and I have attached a spreadsheet that I need to analyze

It contains 72 stock trades and I need to chart the Average "Trade Grade" (column K) but for each chunk/cluster of 10 and another chart for 20 records

So for example using chunks of 10 rows there would be a chart with 8 trade grades (7 chunks of 10, 1 chunk of 2 = 72 trades split in 8 chunks)

And using chunks or cluster of 20 records there would be 4 trade grades   (3 chunks of 20, 1 chunk of 12)

The criteria for splitting the data would be the Entry Date (column E)

The .xls keeps being updated, next week it might have another 15 trades added and the information would need to be clustered differently because now there would be 87 trades instead of 72

Any pointer in the right direction would be helpful

Thanks and Regards,

Isaac

• ###### 1. Re: How to analyze spreadsheet on chunks/clusters of 10 and 20 records

Is there any definition on the range of entry dates to be included in a group/cluster. I don't see any defined range in the sheet you provide. They are clustered manually.

• ###### 2. Re: How to analyze spreadsheet on chunks/clusters of 10 and 20 records

Hello,

That visual clustering I did it manually just to illustrate how they need to be grouped

But the cluster calculation is dynamic, there is no column that indicates which row belongs to which cluster, the criteria would be the following:

Cluster 1 =    10 most recent trades by "Entry Date"

Custer 2 =      next 10 most recent trades not included in Cluster 1

Cluster 3 =     next 10 most recent trades not included in Cluster 1 or 2

...

Cluster X =    the remaining trades not included in any previous cluster

The same logic applies for clusters of 20

As new records get added the rows in the each cluster will change

Regards,

Isaac

• ###### 3. Re: How to analyze spreadsheet on chunks/clusters of 10 and 20 records

Hi Isaac,

Do you know if there is any unique identifier for each row ?

Thanks,

Mani

• ###### 4. Re: How to analyze spreadsheet on chunks/clusters of 10 and 20 records

Hi Mani,

Currently there isn't, but if it helps I could add some consecutive ID for each row to identify each trade or something like that

Regards,

Isaac