mlindauer/aslibscenario

aslibscenario's Issues

ADD check that all algorithms have indeed available performance data

If an algorithm is listed in description.txt, it should also have data in algorithm_runs.arff

TypeError: 'generator' object is not subscriptable

Dear ｍlindauer,

In the execution, I got the following error:

if arff_dict["attributes"][0][0].upper() != "INSTANCE_ID":
TypeError: 'generator' object is not subscriptable

Best

generator' object is not subscriptable

Hello. First, thank you for the nice code ! I'm trying to read one of the aslib datasets calling the method read_scenario from the ASlibScenario class. It starts reading OK but at some point it seems that an error is triggered (tested on python 3.7.4 and with pypy 7.2 on linux) :

CODE:
sdn = ASlibScenario()
sdn.read_scenario('/home/haroldo/inst/algsel/aslib_data/SAT03-16_INDU/')

...
OUTPUT:
Optional file not found: /home/haroldo/inst/algsel/aslib_data/SAT03-16_INDU/ground_truth.arff

Warning (from warnings module):
File "/home/haroldo/dev/ASlibScenario/aslib_scenario/aslib_scenario.py", line 235
description = yaml.load(fh)
YAMLLoadWarning: calling yaml.load() without Loader=... is deprecated, as the default Loader is unsafe. Please read https://msg.pyyaml.org/load for full details.
Have not found algorithm_cutoff_memory
Have not found features_stochastic
Traceback (most recent call last):
File "<pyshell#5>", line 1, in
sdn.read_scenario('/home/haroldo/inst/algsel/aslib_data/SAT03-16_INDU/')
File "/home/haroldo/dev/ASlibScenario/aslib_scenario/aslib_scenario.py", line 187, in read_scenario
self.read_files()
File "/home/haroldo/dev/ASlibScenario/aslib_scenario/aslib_scenario.py", line 225, in read_files
read_func(fn)
File "/home/haroldo/dev/ASlibScenario/aslib_scenario/aslib_scenario.py", line 371, in read_algorithm_runs
if arff_dict["attributes"][0][0].upper() != "INSTANCE_ID":
TypeError: 'generator' object is not subscriptable

sdn.read_scenario('/home/haroldo/inst/algsel/aslib_data/SAT03-16_INDU/')
Optional file not found: /home/haroldo/inst/algsel/aslib_data/SAT03-16_INDU/ground_truth.arff
Have not found algorithm_cutoff_memory
Have not found features_stochastic
Traceback (most recent call last):
File "<pyshell#6>", line 1, in
sdn.read_scenario('/home/haroldo/inst/algsel/aslib_data/SAT03-16_INDU/')
File "/home/haroldo/dev/ASlibScenario/aslib_scenario/aslib_scenario.py", line 187, in read_scenario
self.read_files()
File "/home/haroldo/dev/ASlibScenario/aslib_scenario/aslib_scenario.py", line 225, in read_files
read_func(fn)
File "/home/haroldo/dev/ASlibScenario/aslib_scenario/aslib_scenario.py", line 371, in read_algorithm_runs
if arff_dict["attributes"][0][0].upper() != "INSTANCE_ID":
TypeError: 'generator' object is not subscriptable

Import KFold errors

This repository was automatically installed by installing Autofolio. In the script "aslib_scenario.py" the following is included:
from sklearn.cross_validation import KFold
However, KFold has been moved from the module "cross_validation" to "model_selection" since sklearn 18.2.

sklearn version requirement is not defined in the requirements.txt of autofolio.

Replace inf in feature values