PKDD dataset

[Former Member]

Hi,

I am reading this tutorial:

SAP Predictive Analytics 2.5

2016-03-11

Data Manipulation Scenario

Data Manager User Guide

In this tutorial,  dataset is used as an example to explain concepts. It is called PKDD. I know PKDD is a name of conference, but I think PKDD in the context means something else. Do you know where we can downlaod this dataset?

Thanks,

Rasoul

[achab · Accepted solution]

Hi Karimi,

Can you please explore the great post here: Hands-On Tutorial SAP Predictive Analytics, Aut... | SCN?

Thanks & regards

Antoine

[Former Member · Accepted solution]

Hi Antonie,

I just checked the link you sent me. That is it! The data is available there.

Thanks a lot for your help!

Rasoul

[abdel_dadouche]

Hi,

The reference is to the "PKDD'99 Discovery Challenge - Financial Data Set" available at the following link: http://lisp.vse.cz/pkdd99/Challenge/berka.htm.

Hope this will meet your need.

And thanks for using our Tutorials. Feel free to use the "Like" and share it across you community.

Regards

@bdel

[Former Member]

Hi Abdel,

Thanks a lot for your quick response! It seems that it is no longer possible to download this data set here: http://lisp.vse.cz/pkdd99/

do you know if there is other possibility?

Thanks,

[abdel_dadouche]

Hi,

Have you rtied the link on the following page : http://lisp.vse.cz/pkdd99/Challenge/chall.htm.

Tthe link should be http://lisp.vse.cz/pkdd99/DATA/data_berka.zip.

Regards

Abdel

[Former Member]

Thanks a lot!

I have a problem to import data to hana. In this dataset, there is a table called CARD and in this table there is a column called "issued". It contains the issue data of the card. When the data is imported, by default the type of this column is VARCHAR. If I leave it like that, then I will have problem in Predictive Analytics: To create a PI, I need a column with timestamp type. So, I have to changed the type of this column while importing data. If I do so, this error appears:

Message :

    Batch from record 2 to 893 failed: Timestamp format must be yyyy-mm-dd hh:mm:ss[.fffffffff]: Timestamp format must be yyyy-mm-dd hh:mm:ss[.fffffffff]

java.lang.IllegalArgumentException: Timestamp format must be yyyy-mm-dd hh:mm:ss[.fffffffff]

I know there are different formats to show a timestamp. But if this format is wrong, why it is like this in the original file? and how can I change the format of time stamp?

If you have any idea, it would be great if you tell me.

Best regards,

Rasoul

[achab]

Hi, please post the question in a dedicated thread.

Thanks

Antoine

[abdel_dadouche]

Hi,

are trying to import the data to HANA using the following linkhttp://lisp.vse.cz/pkdd99/DATA/data_berka.zip?

if so please the HANA Delivery unit provided by Andreas Forster available at : https://github.com/AndreasForster/Predictive/raw/master/DataManagerTutorialData.zip

Regards

@bdel

[Former Member]

Hi,

I downloaded the zip file from the link you sent me (https://github.com/AndreasForster/Predictive/raw/master/DataManagerTutorialData.zip)

For each table there is a folder. In the folder there is sql file to create the table, but where are data and how to import them?

Thanks,

Rasoul

[abdel_dadouche]

Hi,

you have to use the HANA Studio Import feature. Please check the following SCN entry . It will import both the structure and the data into your instance.

Best Regards

@bdel

[AndreasForster]

Hi Karimi,

Here is more information on that delivery unit. It belongs to a separate Data Manager tutorial, which is very much based on the same dataset.

https://scn.sap.com/docs/DOC-68745

Many Greetings

Andreas

[Former Member]

Hi Andreas,

The tutorial that I found in this link is much better than the first tutorial that I was reading ( SAP Predictive Analytics 2.5 2016-03-11 Data Manipulation Scenario Data Manager User Guide).

Thanks a lot!

Rasoul

[AndreasForster]

Hello Rasoul,

Glad to hear you like the tutorial!

There are a few other similar ones on classification and time series forecasting. Just in case you are interested.

http://scn.sap.com/docs/DOC-70069

Many Greetings

Andreas