Jerome Pineau's dashboard

Jerome Pineau replied on October 17, 2009 19:19 to the question "click ere" in GoodData:

Sorry Ranchhod - can you be a little more specific? :)
Thanks
J.

Jerome Pineau replied on October 13, 2009 22:04 to the question "Countable/non-countable should be mutually exclusive" in GoodData:

Thanks Tom, that makes sense to me.

Jerome Pineau replied on October 13, 2009 00:53 to the question "How do you join files on PK/FK?" in GoodData:

Thanks Ray. I take it you're not in Prague where it's almost 3am right? :)

Jerome Pineau replied on October 13, 2009 00:42 to the question "How do you join files on PK/FK?" in GoodData:

Ok got it. It works just nice now. It would be really nice if I could mark all columns as don't import with one click for the purpose of loading dimension tables (also numbering the cols would be cool as in 1 of N, 2 of N, 3 of N etc..).

Now if I want to upload an incremental OR if I want to upload the same table (but this time I want to include a col i previously did NOT import) do I have to go thru the whole process again?

Thanks man. This is cool.

Jerome Pineau replied on October 13, 2009 00:19 to the question "How do you join files on PK/FK?" in GoodData:

Thanks Ray,
So you're saying (taking my example again) I have sales_facts, products, and stores. So I should upload say products (no facts), make the connection using product_id, then I should upload stores, make the connection using store_id, and THEN only upload sales_facts. Once all the sets are uploaded THEN I should be able to connect sales_facts to products and stores using those 2 columns?

Jerome Pineau replied on October 10, 2009 19:42 to the question "How do you join files on PK/FK?" in GoodData:

Mind you I am watching

http://www.gooddata.com/good-data-tv/...

so hopefully can figure it out from that.
Thanks.

Jerome Pineau replied on October 10, 2009 19:39 to the question "How do you join files on PK/FK?" in GoodData:

Ok sorry but this time I got the attached error. Here's the DDL for both files:

CREATE TABLE promotion(
promotion_id int ,
promotion_district_id int ,
promotion_name char varying(255) ,
media_type char varying(255) ,
cost decimal(12,2) ,
start_date date ,
end_date date
) from "c:\projects\xsprada\mondrian\promotion.csv"

CREATE TABLE sales_fact_1997(
product_id int ,
time_id int ,
customer_id int ,
promotion_id int ,
store_id int ,
store_sales decimal(12,2) ,
store_cost decimal(12,2) ,
unit_sales decimal(12,2)
) from "c:\projects\xsprada\mondrian\sales_fact_1997.csv"

(don't worry about the FROM part, not relevant) - So my assumption was that I could upload both files and then tell GD ok, now join them on promotion_id. Actually, I thought I could upload multiple dimension files and then tell it join store, promotions and product on corresponding keys. I am still not clear on how to do this and I guess I dont really understand your suggestion then I'm sorry but I am somewhat confused.

Jerome Pineau shared an idea in GoodData on October 10, 2009 19:02:

Quick browsing of existing data files.

Not sure if currently possible but IMO it would be nice to be able to quickly browse say the first N rows of an uploaded data file. (kinda like what displays in the loader form).

Jerome Pineau replied on October 10, 2009 18:53 to the idea "Usability annoyance regarding data uploads" in GoodData:

Actually what I'd like to see is the ability to upload a file and its corresponding DDL as well so i dont have to manually enter all the column info. Even being able to paste DDL somewhere would be nice.

Jerome Pineau asked a question in GoodData on October 10, 2009 18:46:

How do you join files on PK/FK?

I am trying to join 2 CSV datasets from FoodMart namely sales_fact_1997 and promotion - I uploaded both files, defined cols etc..Now I want to join on promotion_id. I'll be damned if I can get it to work. It tells me both files are joinable but in the popup dialog the right had side is always grayed out so I can't complete the join. What am I doing wrong?
Adding a screenshot here:

Jerome Pineau replied on October 10, 2009 18:43 to the question "Countable/non-countable should be mutually exclusive" in GoodData:

Thanks Pavel. I understand the concept but still feel it's confusing to a 1st time user. Maybe it's semantics.
'preciate the help.

Jerome Pineau asked a question in GoodData on October 10, 2009 18:17:

Countable/non-countable should be mutually exclusive

When you define columns on upload, why are the checkboxes "Countable" and "Non-countable" not mutually exclusive?

Jerome Pineau asked a question in GoodData on October 10, 2009 18:04:

What does the "Sorry, only one column can be mapped to each dimension." error msg mean?

When I load a file and map the columns, I sometimes get a "Sorry, only one column can be mapped to each dimension." message - This is confusing. Not sure what it means.

Jerome Pineau asked a question in GoodData on October 10, 2009 15:26:

Changing column header names

To change spelling on a column header you have to double-click the header. This is not intuitive (should be at least some hint somewhere).