search

mdbloice / Pyrea

Multi-view clustering with flexible ensemble structures.
https://pyrea.readthedocs.io
MIT License
21 stars 3 forks source link

TypeError: np.matrix is not supported. Please convert to a numpy array with np.asarray. #2

Open PengleiGao opened 4 months ago

PengleiGao commented 4 months ago

Hi, when I use the package, I have the following issues: code is: d1 = np.random.rand(100,10) d2 = np.random.rand(100,10) d3 = np.random.rand(100,10) data = [d1,d2,d3] params = pyrea.parea_2_genetic(data, k_min=2, k_max=5)

it returns the error: TypeError Traceback (most recent call last) Cell In[14], line 1 ----> 1 params_hierarchical = pyrea.parea_1_genetic(data, k_min=2, k_max=5, k_final=3, n_generations=10, n_population=10)

File ~/anaconda3/envs/eval/lib/python3.10/site-packages/pyrea/core.py:833, in parea_1_genetic(data, k_min, k_max, k_final, n_population, n_generations) 830 stats.register("max", np.max) 832 # Run the genetic algorithm --> 833 pop, log = algorithms.eaSimple(population, toolbox, cxpb=0.7, mutpb=0.2, ngen=n_generations, stats=stats, halloffame=hall_of_fame, verbose=True) 835 print(f"\nSummary:\n{log}") 837 return hall_of_fame[0]

File ~/anaconda3/envs/eval/lib/python3.10/site-packages/deap/algorithms.py:151, in eaSimple(population, toolbox, cxpb, mutpb, ngen, stats, halloffame, verbose) 149 invalid_ind = [ind for ind in population if not ind.fitness.valid] 150 fitnesses = toolbox.map(toolbox.evaluate, invalid_ind) --> 151 for ind, fit in zip(invalid_ind, fitnesses): 152 ind.fitness.values = fit 154 if halloffame is not None:

File ~/anaconda3/envs/eval/lib/python3.10/site-packages/pyrea/core.py:795, in parea_1_genetic..evaluate(individual) 792 with warnings.catch_warnings(): 793 warnings.simplefilter("ignore") --> 795 sil = parea_1(data, 796 individual[0], 797 individual[1], ... 839 xp, is_array_api_compliant = get_namespace(array) 841 # store reference to original array to check if copy is needed when 842 # function returns TypeError: np.matrix is not supported. Please convert to a numpy array with np.asarray.

Do you know how to fix it? Thanks.

mdbloice commented 4 months ago

Hi @PengleiGao

yeah this is because I used the numpy.matrix class and this now deprecated in newer versions of NumPy. I would need to change Pyrea's code to use the numpy.array class instead. Which is on the agenda, and I will do so in the coming days actually. Once I have made the changes, I will update the package and let you know.