gdalessi/openmore

Python libraries (NumPy-based) to perform model order reduction, clustering and data analysis for combustion and aerothermochemistry data

openmore's Issues

The system cannot find the path specified

hank you for your code, it's very useful to perform local PCA and easy to use 🙂 However, I have a problem in the code below and I don't know how to fix it.
My code is simply finding the best clusters for each number of eigenvectors (from 1 to the initial space). Then, I perform local pca on my different clusters and finally evaluate the reconstruction of my initial dataset. Here is my code :

import OpenMORe.clustering as clustering
n_clusters = 4
r2final=[]
r2clusters=[]

for q in range(1,test.shape[1]+1): #number of pc
settings_clustering = {
#centering and scaling options
"center" : True,
"centering_method" : "mean",
"scale" : True,
"scaling_method" : "auto",

#set the initialization method (random, observations, kmeans, pkcia, uniform)
"initialization_method"     : "random",

#set the number of clusters and PCs in each cluster
"number_of_clusters"        : n_clusters,
"number_of_eigenvectors"    : q,

#enable additional options:
"correction_factor"         : "off",    # --> enable eventual corrective coefficients for the LPCA algorithm:
                                        #     'off', 'c_range', 'uncorrelation', 'local_variance', 'phc_multi', 'local_skewness' are available

"classify"                  : False,    # --> call the method to classify a new matrix Y on the basis of the lpca clustering
"write_on_txt"              : False,    # --> write the idx vector containing the label for each observation
"evaluate_clustering"       : True,     # --> enable the calculation of indeces to evaluate the goodness of the clustering
"neighbors_number"          : 0,
}

# First we create the model and then we fit it to the dataset
model = clustering.lpca(test, settings_clustering)
labels_vqpca = model.fit()

r2final=[]
for k in range(n_clusters):
    mask = (labels_vqpca  == k)
    X_k = test[mask] # the mask is used to select the clusters with a specific label
    cnt_k = np.mean(X_k, axis=0)
    scl_k = np.std(X_k, axis=0)
    X0_k = (X_k - cnt_k)/scl_k # we need to center the data in each cluster
    pca.fit(X0_k)
    A_k = pca.components_.T
    Z_k = X0_k @ A_k
    X0_k_rec = Z_k[:,:q] @ A_k[:,:q].T
    X_k_rec = scl_k * X0_k_rec + cnt_k
    r2list = []
    for i in range(X_k.shape[1]): #reconstruction for each elements
        r2 = r2_score(X_k[:,i], X_k_rec[:,i])
        r2list.append(r2)
    r2final.append(r2list)
averages = [sum(elements) / len(elements) for elements in zip(*r2final)]  
r2clusters.append(averages)

However, I get this error which i don't really understand when calling model.fit():

FileNotFoundError: [WinError 3] The system cannot find the path specified: 'Clustering LPCA - 2023_05_05-144151'

Thank you for your help.