cancel
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for 
Search instead for 
Did you mean: 
zdc
‎06-13-2023
  • 9 Posts
  • 9 Likes Given
  • 0 Solutions
  • 0 Likes Received

Re: Why proc gradboost autotune gives different best parameters every ...

by Fluorite | Level 6 in SAS Data Science
‎09-16-2021 03:51 AM
‎09-16-2021 03:51 AM
Hi @sbxkoenk,   Thank you for your reply. I was also expecting differences but not that much. When I augmented the number of trees and put the nthreads option to 1, the differences have been more acceptable. This provided some insight, thank you! ... View more

Re: Why proc gradboost autotune gives different best parameters every ...

by Fluorite | Level 6 in SAS Data Science
‎09-14-2021 02:41 AM
‎09-14-2021 02:41 AM
Hi, Thank you for your reply. Since I use the same data, even if they change, I expect the parameters and the resulting mape value to not change much. But I'm getting quite different results. I can't understand why. ... View more

Re: Why proc gradboost autotune gives different best parameters every ...

by Fluorite | Level 6 in SAS Data Science
‎09-14-2021 02:33 AM
‎09-14-2021 02:33 AM
Hi, Thank you for your reply. I pay special attention to the fact that the data does not change and I use exactly the same data. I run the model again with the best parameters and calculate the mape value. Because of different best parameters every time, it returns different mape values (quite different). Even when I run it over and over with the same best parameters, it returns different results. When I run it in python with the same data, the best parameters values do not change. As a result, the mape value does not change either. I also expect to get the same results every time, but I don't understand why it that much changed. ... View more

Re: Why proc gradboost autotune gives different best parameters every ...

by Fluorite | Level 6 in SAS Data Science
‎09-14-2021 02:22 AM
‎09-14-2021 02:22 AM
Hi,   Thank you for your reply. I had to rerun the code and I run the exact same code each time with exactly the same data (including sorting).   I used default autotune values.    ods noproctitle; proc gradboost data=CASUSER.TRAIN_DATA; target Number / level=interval; input A B C D E F G H J K/ level=interval; autotune tuningparameters=(ntrees samplingrate vars_to_try(init=100) learningrate lasso ridge) objective=mae kfold=5; ods output FitStatistics=Work._Gradboost_FitStats_ VariableImportance=Work._Gradboost_VarImp_; score out=casuser.scored_train copyvars=(_all_); savestate rstore=casuser.model_train; id _all_; run; ... View more

Why proc gradboost autotune gives different best parameters every time...

by Fluorite | Level 6 in SAS Data Science
‎09-13-2021 08:07 AM
‎09-13-2021 08:07 AM
Hi,   I'm using gradient boosting as algorithm for my model in SAS Code. I'm also trying to find the best parameters using autotune. But every time I run the program it returns different best parameters. I couldn't understand the working logic. Isn't there only one best parameter? Why does it return different best parameters each time -unlike python- ? I would be very happy if anyone has come across and knows that problem.   Thank you.  ... View more

SAS sum of next N rows for every row by 2 columns

by Fluorite | Level 6 in SAS Programming
‎08-31-2021 10:08 AM
‎08-31-2021 10:08 AM
Hi, I recently asked a question about SAS sum of next N rows for every row here .   This is exactly the answer to the question and works well (by one column). data want(drop = i n); set have curobs = c nobs = nobs; Sum_Next_6Numbers = 0; do p = c + 1 to 6 + c; if p > nobs then do; Sum_Next_6Numbers = .; leave; end; set have(keep = Number ID rename = (Number = n id = i)) point = p; if id ne i then do; Sum_Next_6Numbers = .; leave; end; Sum_Next_6Numbers + n; end; run;   So basically, I want to sum up next 6 row's values(number column) by 2 columns(id and id2). If there is no 6 months left then this values should be Nan.     I have a dataset like this for each ID and ID2;   data have; input Months ID ID2 Number; 2018-07-01 1 1 0 2018-08-01 1 1 0 2018-09-01 1 1 0 2018-10-01 1 1 1 2018-11-01 1 1 0 2018-12-01 1 1 1 2019-01-01 1 1 0 2019-02-01 1 1 0 2019-03-01 1 1 0 2019-04-01 1 1 0 2019-05-01 1 1 0 2019-06-01 1 1 0 2019-07-01 1 1 0 2019-08-01 1 1 0 2019-09-01 1 1 0 2019-10-01 1 1 0 2019-11-01 1 1 0 2019-12-01 1 1 0 2020-01-01 1 1 0 2020-02-01 1 1 0 2020-03-01 1 1 0 2020-04-01 1 1 0 2020-05-01 1 1 0 2020-06-01 1 1 0 2020-07-01 1 1 0 2020-08-01 1 1 0 2020-09-01 1 1 0 2020-10-01 1 1 0 2020-11-01 1 1 1 2020-12-01 1 1 0 2021-01-01 1 1 0 2021-02-01 1 1 0 2021-03-01 1 1 0 2021-04-01 1 1 0 2018-07-01 1 2 0 2018-08-01 1 2 0 2018-09-01 1 2 0 2018-10-01 1 2 0 2018-11-01 1 2 0 2018-12-01 1 2 0 2019-01-01 1 2 0 2019-02-01 1 2 0 2019-03-01 1 2 0 2019-04-01 1 2 0 2019-05-01 1 2 0 2019-06-01 1 2 0 2019-07-01 1 2 0 2019-08-01 1 2 0 2019-09-01 1 2 0 2019-10-01 1 2 0 2019-11-01 1 2 0 2019-12-01 1 2 0 2020-01-01 1 2 0 2020-02-01 1 2 0 2020-03-01 1 2 0 2020-04-01 1 2 0 2020-05-01 1 2 0 2020-06-01 1 2 0 2020-07-01 1 2 0 2020-08-01 1 2 0 2020-09-01 1 2 0 2020-10-01 1 2 0 2020-11-01 1 2 0 2020-12-01 1 2 0 2021-01-01 1 2 0 2021-02-01 1 2 0 2021-03-01 1 2 0 2021-04-01 1 2 0 2018-07-01 2 1 0 2018-08-01 2 1 0 2018-09-01 2 1 0 2018-10-01 2 1 0 2018-11-01 2 1 0 2018-12-01 2 1 0 2019-01-01 2 1 0 2019-02-01 2 1 0 2019-03-01 2 1 0 2019-04-01 2 1 0 2019-05-01 2 1 0 2019-06-01 2 1 0 2019-07-01 2 1 0 2019-08-01 2 1 0 2019-09-01 2 1 0 2019-10-01 2 1 0 2019-11-01 2 1 0 2019-12-01 2 1 0 2020-01-01 2 1 0 2020-02-01 2 1 0 2020-03-01 2 1 0 2020-04-01 2 1 0 2020-05-01 2 1 0 2020-06-01 2 1 0 2020-07-01 2 1 0 2020-08-01 2 1 1 2020-09-01 2 1 0 2020-10-01 2 1 0 2020-11-01 2 1 0 2020-12-01 2 1 0 2021-01-01 2 1 0 2021-02-01 2 1 0 2021-03-01 2 1 0 2021-04-01 2 1 0 ....... ....... .......   I want a dataset like this;   data want; input Months ID ID2 Number Sum_Next_6Numbers; 2018-07-01 1 1 0 2 2018-08-01 1 1 0 2 2018-09-01 1 1 0 2 2018-10-01 1 1 1 1 2018-11-01 1 1 0 1 2018-12-01 1 1 1 0 2019-01-01 1 1 0 0 2019-02-01 1 1 0 0 2019-03-01 1 1 0 0 2019-04-01 1 1 0 0 2019-05-01 1 1 0 0 2019-06-01 1 1 0 0 2019-07-01 1 1 0 0 2019-08-01 1 1 0 0 2019-09-01 1 1 0 0 2019-10-01 1 1 0 0 2019-11-01 1 1 0 0 2019-12-01 1 1 0 0 2020-01-01 1 1 0 0 2020-02-01 1 1 0 0 2020-03-01 1 1 0 0 2020-04-01 1 1 0 0 2020-05-01 1 1 0 1 2020-06-01 1 1 0 1 2020-07-01 1 1 0 1 2020-08-01 1 1 0 1 2020-09-01 1 1 0 1 2020-10-01 1 1 0 1 2020-11-01 1 1 1 NaN 2020-12-01 1 1 0 NaN 2021-01-01 1 1 0 NaN 2021-02-01 1 1 0 NaN 2021-03-01 1 1 0 NaN 2021-04-01 1 1 0 NaN 2018-07-01 1 2 0 0 2018-08-01 1 2 0 0 2018-09-01 1 2 0 0 2018-10-01 1 2 0 0 2018-11-01 1 2 0 0 2018-12-01 1 2 0 0 2019-01-01 1 2 0 0 2019-02-01 1 2 0 0 2019-03-01 1 2 0 0 2019-04-01 1 2 0 0 2019-05-01 1 2 0 0 2019-06-01 1 2 0 0 2019-07-01 1 2 0 0 2019-08-01 1 2 0 0 2019-09-01 1 2 0 0 2019-10-01 1 2 0 0 2019-11-01 1 2 0 0 2019-12-01 1 2 0 0 2020-01-01 1 2 0 0 2020-02-01 1 2 0 0 2020-03-01 1 2 0 0 2020-04-01 1 2 0 0 2020-05-01 1 2 0 0 2020-06-01 1 2 0 0 2020-07-01 1 2 0 0 2020-08-01 1 2 0 0 2020-09-01 1 2 0 0 2020-10-01 1 2 0 0 2020-11-01 1 2 0 NaN 2020-12-01 1 2 0 NaN 2021-01-01 1 2 0 NaN 2021-02-01 1 2 0 NaN 2021-03-01 1 2 0 NaN 2021-04-01 1 2 0 NaN 2018-07-01 2 1 0 0 2018-08-01 2 1 0 0 2018-09-01 2 1 0 0 2018-10-01 2 1 0 0 2018-11-01 2 1 0 0 2018-12-01 2 1 0 0 2019-01-01 2 1 0 0 2019-02-01 2 1 0 0 2019-03-01 2 1 0 0 2019-04-01 2 1 0 0 2019-05-01 2 1 0 0 2019-06-01 2 1 0 0 2019-07-01 2 1 0 0 2019-08-01 2 1 0 0 2019-09-01 2 1 0 0 2019-10-01 2 1 0 0 2019-11-01 2 1 0 0 2019-12-01 2 1 0 0 2020-01-01 2 1 0 0 2020-02-01 2 1 0 1 2020-03-01 2 1 0 1 2020-04-01 2 1 0 1 2020-05-01 2 1 0 1 2020-06-01 2 1 0 1 2020-07-01 2 1 0 1 2020-08-01 2 1 1 0 2020-09-01 2 1 0 0 2020-10-01 2 1 0 0 2020-11-01 2 1 0 NaN 2020-12-01 2 1 0 NaN 2021-01-01 2 1 0 NaN 2021-02-01 2 1 0 NaN 2021-03-01 2 1 0 NaN 2021-04-01 2 1 0 NaN ....... ....... ....... I did it with a nested for loop in python. I'm new to sas and I'm confused. Is there a way to do this in SAS? Thanks in advance.      ... View more

Re: SAS sum of next N rows for every row

by Fluorite | Level 6 in SAS Programming
‎08-20-2021 07:40 AM
‎08-20-2021 07:40 AM
Thank you very much, this helped me a lot. I'm appreciated! ... View more

Re: SAS sum of next N rows for every row

by Fluorite | Level 6 in SAS Programming
‎08-20-2021 07:39 AM
‎08-20-2021 07:39 AM
I don't have SAS/ETS and this code works very well. Thank you very much I'm appreciated! ... View more

SAS sum of next N rows for every row

by Fluorite | Level 6 in SAS Programming
‎08-20-2021 03:15 AM
‎08-20-2021 03:15 AM
I have a dataset like this for each ID;   data have; input Months ID Number; 2018-07-01 1 0 2018-08-01 1 0 2018-09-01 1 1 2018-10-01 1 3 2018-11-01 1 1 2018-12-01 1 2 2019-01-01 1 0 2019-02-01 1 0 2019-03-01 1 1 2019-04-01 1 0 2019-05-01 1 0 2019-06-01 1 0 2019-07-01 1 1 2019-08-01 1 0 2019-09-01 1 0 2019-10-01 1 2 2019-11-01 1 0 2019-12-01 1 0 2020-01-01 1 0 2020-02-01 1 0 2020-03-01 1 0 2020-04-01 1 0 2020-05-01 1 0 2020-06-01 1 0 2020-07-01 1 0 2020-08-01 1 1 2020-09-01 1 0 2020-10-01 1 0 2020-11-01 1 1 2020-12-01 1 0 2021-01-01 1 0 2021-02-01 1 1 2021-03-01 1 1 2021-04-01 1 0 2018-07-01 2 0 ....... ....... ....... (Similar values for each ID)   I want a dataset like this; data want; input Months ID Number Sum_Next_6Numbers; 2018-07-01 1 0 7 2018-08-01 1 0 7 2018-09-01 1 1 7 2018-10-01 1 3 4 2018-11-01 1 1 3 2018-12-01 1 2 1 2019-01-01 1 0 2 2019-02-01 1 0 2 2019-03-01 1 1 1 2019-04-01 1 0 3 2019-05-01 1 0 3 2019-06-01 1 0 3 2019-07-01 1 1 2 2019-08-01 1 0 2 2019-09-01 1 0 2 2019-10-01 1 2 0 2019-11-01 1 0 0 2019-12-01 1 0 0 2020-01-01 1 0 0 2020-02-01 1 0 1 2020-03-01 1 0 1 2020-04-01 1 0 1 2020-05-01 1 0 2 2020-06-01 1 0 2 2020-07-01 1 0 2 2020-08-01 1 1 2 2020-09-01 1 0 3 2020-10-01 1 0 3 2020-11-01 1 1 Nan 2020-12-01 1 0 Nan 2021-01-01 1 0 Nan 2021-02-01 1 1 Nan 2021-03-01 1 1 Nan 2021-04-01 1 0 Nan 2018-07-01 2 0 0 ....... ....... ....... I want to sum up next 6 row's values(for number column). If there is no 6 months left then this values should be Nan.   e. g. for first row(2018-07-01) want to sum next 6 rows (2018-08-01,2018-09-01,2018-10-01,2018-11-01,2018-12-01,2019-01-01) of Number column.   Is there a way to do this? Thank you in advance. ... View more
Contact Me
Online Status
Offline
Date Last Visited
‎06-13-2023 12:48 PM
Likes given to
Quick links