sigmoid range constrain about xgboost HOT 8 CLOSED

hitting this condition means you set base_score outside (0,1).

from xgboost.

I've just executed the demo script as it is... Am I supposed to change anything in it? Thank you

from xgboost.

This is strange, the code can run directly. please make sure you have the
most recent version in GitHub

On Tuesday, July 8, 2014, giuliohome [email protected] wrote:

I've just executed the demo script as it is... Am I supposed to change
anything in it? Thank you

—
Reply to this email directly or view it on GitHub
https://github.com/tqchen/xgboost/issues/14#issuecomment-48288947.

Sincerely,

Tianqi Chen
Computer Science & Engineering, University of Washington

from xgboost.

sure, yesterday I downloaded xgboost from here:
https://github.com/tqchen/xgboost/archive/master.zip

That is strange indeed: base_score appears to be nan when the assertion is checked...
any other clues?

from xgboost.

This is strange, the code can run directly. please make sure you have the
most recent version in GitHub .

On Tuesday, July 8, 2014, giuliohome <[email protected]
javascript:_e(%7B%7D,'cvml','[email protected]');> wrote:

I've just executed the demo script as it is... Am I supposed to change
anything in it? Thank you

—
Reply to this email directly or view it on GitHub
https://github.com/tqchen/xgboost/issues/14#issuecomment-48288947.

Sincerely,

Tianqi Chen
Computer Science & Engineering, University of Washington

from xgboost.

I'm sure that I have the most recent version, because this is what I've done again also today:
git clone https://github.com/tqchen/xgboost.git

then I did 'make' from the xgboost folder and 'make' from the python folder.

Finally I went to the kaggle demo folder, where I downloaded the 2 csv data files from kaggle.com and I executed the run.sh script. The output follows:

finish loading from csv
weight statistics: wpos=1522.37, wneg=904200, ratio=593.94
loading data end, start to boost trees
Error:sigmoid range constrain

ERROR in higgs-numpy.py

from xgboost.

This is my result on totally new machine

[@eos1 kaggle-higgs]$ ./run.sh 
finish loading from csv 
weight statistics: wpos=1522.37, wneg=904200, ratio=593.94
loading data end, start to boost trees
...
[119]   train-auc:0.945876      [email protected]:5.935486
finish training
finish loading from csv 
finished writing into prediction file

Can you try run another machine? What is your system in detail?

from xgboost.

Yes, you're right. Moded to another server, where it is working.
Thanks, issue closed.

from xgboost.