fix: numpy.random.RandomState is no longer an accepted random state b…

…y hyperopt. Use numpy.random.Generator instead (#50)

auto_surprise/context/result_logging_manager.py

@@ -1,6 +1,5 @@
 import time
 import datetime
-import pathlib
 import logging
 
 class ResultLoggingManager:

auto_surprise/engine.py

@@ -24,7 +24,7 @@
 
 class Engine(object):
     def __init__(
-        self, verbose=True, algorithms=FULL_ALGO_LIST, random_state=numpy.random
+        self, verbose=True, algorithms=FULL_ALGO_LIST, random_state=numpy.random.default_rng()
     ):
         """
         Initialize new engine

docs/usage/reproducing_experiments.rst

@@ -7,13 +7,11 @@ Reproducing Experiments
 You may want to make sure that your results are reproducible. This can be done easily by setting the seed and random state when initializing `Engine`.
 
 .. code-block:: python
-
+    import numpy
     from auto_surprise.engine import Engine
 
-    random.seed(123)
-    numpy.random.seed(123)
-
     # Intitialize auto surprise engine with random state set
-    engine = Engine(verbose=True, random_state=numpy.random.RandomState(123))
+    engine = Engine(verbose=True, random_state=numpy.random.default_rng(12345))
 
 This will make sure that you're results will be exactly the same, provided you're other training params also stay the same.
+

requirements.txt

@@ -1,19 +1,22 @@
-cloudpickle==2.0.0
-commonmark==0.9.1
-future==0.18.2
+cloudpickle==2.2.1
+future==0.18.3
 hyperopt==0.2.7
-joblib==1.1.0
-networkx==2.7.1
-numpy==1.22.3
-pandas==1.4.1
-py4j==0.10.9.5
-Pygments==2.11.2
+joblib==1.2.0
+markdown-it-py==2.2.0
+mdurl==0.1.2
+networkx==3.1
+numpy==1.24.2
+pandas==2.0.0
+py4j==0.10.9.7
+Pygments==2.14.0
 python-dateutil==2.8.2
-pytz==2022.1
-rich==12.0.1
-scikit-learn==1.0.2
-scikit-surprise==1.1.1
-scipy==1.8.0
+pytz==2023.3
+rich==13.3.3
+scikit-learn==1.2.2
+scikit-surprise==1.1.3
+scipy==1.10.1
 six==1.16.0
 threadpoolctl==3.1.0
-tqdm==4.63.1
+tqdm==4.65.0
+typing-extensions==4.5.0
+tzdata==2023.3