AHRR = -nan when trying model selection
murak038 opened this issue · 3 comments
I was performing model selection using the following code (pulled from the example provided in the UserGuide.ipynb):
import pandas as pd
from SLIM import SLIM, SLIMatrix
traindata = pd.read_csv('~/SLIM/SLIM_web_training.ijv', delimiter = ' ', header=None)
valdata = pd.read_csv('~/SLIM/SLIM_web_test.ijv', delimiter = ' ', header=None)
trainmat = SLIMatrix(traindata)
# initlaize the validation matrix using the training matrix
valmat = SLIMatrix(valdata, trainmat)
params = {'dbglvl':3,
'algo':'cd',
'nthreads':1,
'l1r':1.,
'l2r':1.,
'optTol':1e-7,
'niters':100}
l1s = [0.0001,0.001,0.005,0.01,0.05, 0.1, 0.5, 1, 2, 4,10,20,30,40,50,60,70,80,90,100]
l2s = [0.0001,0.001,0.005,0.01,0.05, 0.1, 0.5, 1, 2,4,5,10,20,30,40,50,60,70,80,90,100]
model = SLIM()
model.mselect(params, trainmat, valmat, l1s, l2s, nrcmds=10)
The dataset is formatted exactly the same as the training and test set provided in the test
folder. An example screenshot is provided below.
When I run model selection on the slim model, I get an output file that has an AHRR of -nan for all of the different hyperparameters that I tested. I provided the output.txt file below. I was wondering why the AHRR is -nan and how the code or data structure can be changed to fix the error and test the model's performance for various hyper parameters.
Data structure:
user_id item_id ratings
Hi Rohit,
Thanks for your feedback. Can you please provide this dataset so that I can reproduce the bug and figure out what is happening here?
Zeren Shui
Here is a .zip file with the training and test set that I used for model selection.
Hi Rohit,
The bug is solved, please try again. Thanks again for your feedback.
Zeren Shui