def main():
clusteringMethod = getClusteringMethod()
filename = getDataset()
if clusteringMethod == "k":
kmeans.main(filename)
elif clusteringMethod == "s":
spectralClustering.main(filename)
def main(path): d = [10, 50, 100, 200, 300, 400, 500, 1000, 2000] l = [] for x in d: print 'Starting kmeans with ' + str(x) +' centroids...' kmeans.main(path, 400) print 'Finished kmeans, proceding with centroidFinder...' centroidFinderP.main(path, x) print 'Finished CF, running kfold10...' a = kfold.main(path, 10) l.append(a) print l
def main(k, n, Random):
"""
The main function of the code
:param k: int , amount of centers
:param n: int, number of points
:param Random: boolean, the Random variable described in the assignment
"""
# import is done here inorder to prevent calling the kmeans module (from tasks.py) before build.
import kmeans
if not Random and (k is None or n is None):
Error('Unable to determine k, n', __file__)
if (k is not None) and (n is not None):
k, n = int(k), int(n)
if k >= n:
Error('k >= n', __file__)
print_max_capacity()
d = random.randint(2, 3)
n, K, X, centers = create_data(n, d, k, Random) # this returned K is the one that was used in the data generation
print(f'The k that was used (K) to create the data is {K}')
print(f'The n that was used to create the data is {n}')
print(f'The d that was used to create the data is {d}')
files.build_data_text_file(X, centers)
"""
Both The Spectral and the Kmeans algorithms end up sending vectors to the kmeans module.
The only difference is which vectors.
Here those vectors (spectral_observations & kmeans_observations) are being computed.
Note: The if else section here is needed because
--> if Random=True, k should be computed by the
eigengap heuristic, and that k will be of use in the following code.
--> else, the actual k is needed when computing U in the spectral module.
"""
if Random:
spectral_observations, k = spectral.main(X)
else:
spectral_observations, _ = spectral.main(X, k) # else => k = K
kmeans_observations = X
spectral_res = kmeans.main(spectral_observations, k, n, k)
kmeans_res = kmeans.main(kmeans_observations, k, n, d)
files.build_clusters_text_file(k, spectral_res, kmeans_res)
files.build_clusters_pdf_file(K, k, n, d, spectral_res, kmeans_res, centers)
def main(argv): argc = len(argv) if (argc is not 5): print '\nWarning: incorrect argument(s) to bagOfWords.py. Expected arguments:\n\n' \ '\t- train_dir (required)\t: directory containing training *.fts files.\n' \ '\t- test_dir (required)\t: directory containing testing *.fts files.\n' \ '\t- N_fts (required)\t: number of *.fts files to use for clustering.\n'\ '\t- K (required)\t\t: number of clusters in K-Means clustering\n' sys.exit(1) train_dir = argv[1] test_dir = argv[2] n_fts = int(argv[3]) n_clusters = int(argv[4]) # 1. Combine files containing SIFT features of individual images into a single file for clustering. print '\n Preparing for clustering...\n' concatFile = concatenateFeatures.main([train_dir, n_fts]) # 2. Cluster features into K clusters with k-means. print '\n Clustering...\n' centersFile = kmeans.main([concatFile, n_clusters]) # 3. Generate histograms for each training file using the cluster centers. print '\n Generating histograms for training files...\n' generateHistogram.main([centersFile, train_dir]) # 4. Generate histograms for each testing file using the cluster centers. print '\n Generating histograms for testing files...\n' generateHistogram.main([centersFile, test_dir]) print '\n Bag of Words done!\n'
def main(argv):
argc = len(argv)
if (argc is not 5):
print '\nWarning: incorrect argument(s) to bagOfWords.py. Expected arguments:\n\n' \
'\t- train_dir (required)\t: directory containing training *.fts files.\n' \
'\t- test_dir (required)\t: directory containing testing *.fts files.\n' \
'\t- N_fts (required)\t: number of *.fts files to use for clustering.\n'\
'\t- K (required)\t\t: number of clusters in K-Means clustering\n'
sys.exit(1)
train_dir = argv[1]
test_dir = argv[2]
n_fts = int(argv[3])
n_clusters = int(argv[4])
# 1. Combine files containing SIFT features of individual images into a single file for clustering.
print '\n Preparing for clustering...\n'
concatFile = concatenateFeatures.main([train_dir, n_fts])
# 2. Cluster features into K clusters with k-means.
print '\n Clustering...\n'
centersFile = kmeans.main([concatFile, n_clusters])
# 3. Generate histograms for each training file using the cluster centers.
print '\n Generating histograms for training files...\n'
generateHistogram.main([centersFile, train_dir])
# 4. Generate histograms for each testing file using the cluster centers.
print '\n Generating histograms for testing files...\n'
generateHistogram.main([centersFile, test_dir])
print '\n Bag of Words done!\n'
def handleButton(self):
# print blah
# print party.blah(1)
# print dir(self.lineEdit)
# if self.wind is None:
strt = tfidfclassification.blah(self.lineEdit.text())
strr = kmeans.main(self.lineEdit.text())
wind = secWin(self)
wind.wind1 = strt
wind.wind2 = strr
self.hide()
wind.show()
def upload_numeric_cluster():
if request.method == 'POST':
# check if the post request has the file part
if 'file' not in request.files:
flash('No file part')
return redirect(request.url)
file = request.files['file']
# if user does not select file, browser also
# submit an empty part without filename
if file.filename == '':
flash('No selected file')
return redirect(request.url)
if file and allowed_file(file.filename):
filename = secure_filename(file.filename)
if filename == "kmean-config.config":
file.save(
os.path.join(
'/home/nexxus/Python/Final_Year/static/files/kmeans',
filename))
CONFIG_kMEAN = '/home/nexxus/Python/Final_Year/static/files/kmeans/' + filename
from kmeans import main
main(CONFIG_kMEAN)
elif filename == "meanshift-config.config":
file.save(
os.path.join(
'/home/nexxus/Python/Final_Year/static/files/meanshift',
filename))
config_meanshift = '/home/nexxus/Python/Final_Year/static/files/meanshift/' + filename
from Meanshift import main
main(config_meanshift)
else:
file.save(
os.path.join('/home/nexxus/Python/Final_Year/static/csv',
filename))
return redirect(request.url)
return render_template('numericclustering.html')
def main(args):
# DP = open("datapoints.txt","w").close()
SP = open("stat_plot.txt","w").close()
ST = open("statistics.txt","w").close()
num_points=0
pointx = ''
while num_points <500000:
num_iter = 0
num_points +=50000
num_generated = 50000
# Create num_points random Points in n-dimensional space
# num_gen, coords, lowerx, upperx, lowery, uppery = int(num_generated*.1), 2,1, 900, 1,900
# pointx = makeRandomPoint(num_gen, lowerx, upperx, lowery, uppery)
# num_gen, coords, lowerx, upperx, lowery, uppery = int(num_generated*.2), 2,50, 350, 50,500
# pointx = makeRandomPoint(num_gen, lowerx, upperx, lowery, uppery)
# num_gen, coords, lowerx, upperx, lowery, uppery = int(num_generated*.2), 2,410, 600,10,550
# pointx = makeRandomPoint(num_gen, lowerx, upperx, lowery, uppery)
#num_gen, coords, lowerx, upperx, lowery, uppery = int(num_generated*.2), 2,600, 890, 600,900
# pointx = makeRandomPoint(num_gen, lowerx, upperx, lowery, uppery)
# num_gen, coords, lowerx, upperx, lowery, uppery = int(num_generated*.1), 2,150, 300,650,900
# pointx = makeRandomPoint(num_gen, lowerx, upperx, lowery, uppery)
# num_gen, coords, lowerx, upperx, lowery, uppery = int(num_generated*.2), 2,650, 880,50,470
# pointx = makeRandomPoint(num_gen, lowerx, upperx, lowery, uppery)
while num_iter <1:
num_iter +=1
filename = "centroid2d_" + `num_iter` +".txt"
shutil.copy(filename,"centroidinput.txt")
datafilename = "datapoints" + `num_points` + ".txt"
#shutil.copy("datapoints1000.txt","datapoints.txt")
# shutil.copy("datapoints.txt", datafilename)
shutil.copy(datafilename,"datapoints.txt")
kmeans.main(num_points)
# final clustering of the datapoints
# os.system("python mapperfinal.py")
#Renaming and finalizing of files
newfilename = "statistics_4n_" + `num_points` +"_"+ `num_iter` +".txt"
os.rename("statistics.txt", newfilename)
# newname = "cluster1_" + `num_points` +".txt"
# os.rename("cluster1.txt", newname)
# newname = "cluster2_" + `num_points` +".txt"
# os.rename("cluster2.txt", newname)
# newname = "cluster3_" + `num_points` +".txt"
# os.rename("cluster3.txt", newname)
# newname = "cluster4_" + `num_points` +".txt"
# os.rename("cluster4.txt", newname)
if os.path.exists("cluster5.txt"):
newname = "cluster5_" + `num_points` +".txt"
os.rename("cluster5.txt", newname)
# newname = "cluster6_" + `num_points` +".txt"
# os.rename("cluster6.txt", newname)
#plot.main()
# DP = open("datapoints.txt","w").close()
# SP = open("stat_plot3d.txt","w").close()
num_points=0
pointx = ''
while num_points <500000:
num_iter = 0
num_points +=50000
num_generated = 50000
# # Create num_points random Points in n-dimensional space
# num_gen, coords, lowerx, upperx, lowery, uppery,lower,upper = int(num_generated*.1), 2,1, 900, 1,900,1,900
# pointx = makeRandom3dPoint(num_gen, lowerx, upperx, lowery, uppery,lower,upper)
# num_gen, coords, lowerx, upperx, lowery, uppery,lower,upper = int(num_generated*.2), 2,50, 250, 50,300,50,400
# pointx = makeRandom3dPoint(num_gen, lowerx, upperx, lowery, uppery,lower,upper)
# num_gen, coords, lowerx, upperx, lowery, uppery,lower,upper = int(num_generated*.2), 2,510, 700,0,250,0,200
# pointx = makeRandom3dPoint(num_gen, lowerx, upperx, lowery, uppery,lower,upper)
# num_gen, coords, lowerx, upperx, lowery, uppery,lower,upper = int(num_generated*.2), 2,0, 200,650,900,600,750
# pointx = makeRandom3dPoint(num_gen, lowerx, upperx, lowery, uppery,lower,upper)
# num_gen, coords, lowerx, upperx, lowery, uppery,lower,upper = int(num_generated*.2), 2,650, 880,50,470,800,900
# pointx = makeRandom3dPoint(num_gen, lowerx, upperx, lowery, uppery,lower,upper)
# num_gen, coords, lowerx, upperx, lowery, uppery,lower,upper = int(num_generated*.1), 2,800, 900,700,900,700,900
# pointx = makeRandom3dPoint(num_gen, lowerx, upperx, lowery, uppery,lower,upper)
while num_iter <1:
num_iter +=1
print num_iter
filename = "centroid3d_" + `num_iter` +".txt"
shutil.copy(filename,"centroidinput.txt")
datafilename = "datapoints3d" + `num_points` + ".txt"
#shutil.copy("datapoints1000.txt","datapoints.txt")
# shutil.copy("datapoints.txt", datafilename)
shutil.copy(datafilename, "datapoints.txt")
kmeans3d.main(num_points)
# final clustering of the datapoints
# os.system("python mapperfinal3d.py")
#Renaming and finalizing of files
newfilename = "statistics3d_4n_" + `num_points` +"_"+ `num_iter` +".txt"
os.rename("statistics3d.txt", newfilename)
# INNER join Peaqock.dbo.Clients ON IdClient=IdPersonne)""",con)
#
#Id_Personne_KYC = list(set(clustered.IdPersonne))
#data = pd.DataFrame()
#for ix in Id_Personne_KYC :
# if data.empty :
# data = df[df['IdPersonne'] == ix]
# else :
# data = pd.concat([df[df['IdPersonne'] == ix],data])
#df = data
df = pd.read_csv('./clients_questionnaire.csv')
df = df.drop('Unnamed: 0', axis=1)
clustered = main()
clustered.index = clustered.IdPersonne
print('Importation des données terminée.', int(time.time() - timee))
##Suppression des variables inutiles ou vides
#list_to_drop = ['Unnamed: 0','IdPersonne','IdCompteEspece','IdOrdreEx','CodeRejet','PrixStop','Marge','QteMinimale','QteDevoilee','NamedOrder','StatusOrdre']
#df = df.drop(list_to_drop,axis = 1)
#Ajout de features
df['TypeValeur'] = df['IdTypeValeur'].replace(all_id_TypeValeurs,
all_label_TypeValeurs)
#%%Extraction des montants Brut pour chaque annee pour chaque client
# Base de données : dbase_montant_brut
import os
import kmeans
if __name__ == '__main__':
for i in range(30):
filenames_eachrun = 'C:/Users/faica/OneDrive/Documents/dev/clustering_od/results.csv'
results_file = 'C:/Users/faica/OneDrive/Documents/dev/clustering_od/results_agg.csv'
kmeans.main()
print()
print('Run :' + str(i))
with open(results_file, "a") as f_end:
with open(filenames_eachrun, "rt") as f:
#reading first line to save the column names
firstline = f.readline()
firstline = firstline.rstrip('\n') + ',run' + '\n'
f_end.write(firstline)
for line in f:
line = line.rstrip('\n') + ',' + str(i) + '\n'
f_end.write(line)