def image_plot(folder):
[ini_coord_x, ini_coord_y, ini_coord_z] = coord_nodes(
pattern_folder=
"C:/Users/mllamosa/Dropbox/2018/Applus/welding patterns all/",
file_name="Initial-Bottom.rpt")
xi, yi = np.meshgrid(np.unique(ini_coord_x), np.unique(ini_coord_z))
name = folder.split(os.sep)[-2] + "_" + folder.split(os.sep)[-1]
files = glob.glob(folder + "/*CH*rpt")
if len(files) > 0:
features = np.zeros(64)
for file in files:
# print(file)
CH = file.split(os.sep)[-1].split("-")[-1].split(".")[0][2:]
d = file.split(os.sep)[-1].split("-")[0].split(".")[0][1:]
# print(CH, d)
features[int(CH) - 1] = int(d)
print("Using distortion at scenario", glob.glob(folder + "/D64*rpt"))
try:
distortions_data = open(glob.glob(folder + "/D64*rpt")[0],
"r").read()
except IndexError:
print(glob.glob(folder + "/D64*rpt"))
distortions_coord_x = np.array(distortions_data[re.search(
"-------------------------------------------------------------------------------------------------",
distortions_data).end():re.search("Minimum", distortions_data).
start()].split()[3:]
[::6]).astype(float)
distortions_coord_y = np.array(distortions_data[re.search(
"-------------------------------------------------------------------------------------------------",
distortions_data).end():re.search("Minimum", distortions_data).
start()].split()[4:]
[::6]).astype(float)
distortions_coord_z = np.array(distortions_data[re.search(
"-------------------------------------------------------------------------------------------------",
distortions_data).end():re.search("Minimum", distortions_data).
start()].split()[5:]
[::6]).astype(float)
zi = griddata((ini_coord_x + distortions_coord_x,
ini_coord_z + distortions_coord_z),
ini_coord_y + distortions_coord_y, (xi, yi),
method='cubic')
plt.imshow(zi, origin='lower', interpolation='none')
plt.savefig('images/' + name + '_img.png')
return name
def read_distortion(pattern_folder=None):
[ini_coord_x, ini_coord_y,
ini_coord_z] = coord_nodes(pattern_folder=pattern_folder,
file_name="Initial-Bottom.rpt")
xyz = {}
for folder_mask in glob.glob(pattern_folder + "/Mask*"):
mask = np.loadtxt(folder_mask + "/mask.tab", delimiter=",")
print(mask.shape)
for folder in glob.glob(folder_mask + "/Scenario*"):
print(folder)
files = glob.glob(folder + "/*CH*rpt")
if len(files) > 0:
features = np.zeros(64)
for file in files:
# print(file)
CH = file.split(
os.sep)[-1].split("-")[-1].split(".")[0][2:]
d = file.split(os.sep)[-1].split("-")[0].split(".")[0][1:]
# print(CH, d)
features[int(CH) - 1] = int(d)
print("Using distortion at scenario",
glob.glob(folder + "/D64*"))
distortions_data = open(
glob.glob(folder + "/D64*rpt")[0], "r").read()
distortions_coord_x = np.array(distortions_data[re.search(
"-------------------------------------------------------------------------------------------------",
distortions_data).end(
):re.search("Minimum", distortions_data).start()].split()
[3:][::6]).astype(float)
distortions_coord_y = np.array(distortions_data[re.search(
"-------------------------------------------------------------------------------------------------",
distortions_data).end(
):re.search("Minimum", distortions_data).start()].split()
[4:][::6]).astype(float)
distortions_coord_z = np.array(distortions_data[re.search(
"-------------------------------------------------------------------------------------------------",
distortions_data).end(
):re.search("Minimum", distortions_data).start()].split()
[5:][::6]).astype(float)
xyz[folder_mask.split(os.sep)[-1] + "_" +
folder.split(os.sep)[-1]] = np.array([
ini_coord_x + distortions_coord_x,
ini_coord_z + distortions_coord_z,
ini_coord_y + distortions_coord_y
])
return xyz
def distortion_to_image(pattern_folder=None):
[ini_coord_x, ini_coord_y,
ini_coord_z] = coord_nodes(pattern_folder=pattern_folder,
file_name="Initial-Bottom.rpt")
xi, yi = np.meshgrid(np.unique(ini_coord_x), np.unique(ini_coord_z))
data = read_distortion(pattern_folder=pattern_folder)
for filename in data.keys():
zi = griddata((data[filename][0], data[filename][1]),
data[filename][2], (xi, yi),
method='cubic')
plt.imshow(zi, origin='lower', interpolation='none')
plt.savefig('images/' + filename + '_img.png')
def pattern_sequence_masked_single(scenario_folder):
[ini_coord_x, ini_coord_y, ini_coord_z] = coord_nodes(
pattern_folder=
"C:/Users/mllamosa/Dropbox/2018/Applus/welding patterns all/",
file_name="Initial-Bottom.rpt")
xi, yi = np.meshgrid(np.unique(ini_coord_x), np.unique(ini_coord_z))
centroids_matrix = np.mgrid[(max(ini_coord_x) - min(ini_coord_x)) /
16:max(ini_coord_x) -
(max(ini_coord_x) - min(ini_coord_x)) / 16:8j,
(max(ini_coord_x) - min(ini_coord_x)) /
16:max(ini_coord_x) -
(max(ini_coord_x) - min(ini_coord_x)) / 16:8j]
centroids_matrix = centroids_matrix.reshape(
2, centroids_matrix.shape[1] * centroids_matrix.shape[2]).T
coord_nodes_chunck = np.array([ini_coord_x, ini_coord_y, ini_coord_z]).T
distances_matrix = distance.cdist(
centroids_matrix,
coord_nodes_chunck.mean(axis=0)[[0, 2]].reshape(1, 2), "euclidean")
distances_matrix_corner = distance.cdist(
centroids_matrix, np.array([[min(ini_coord_x),
min(ini_coord_x)]]), "euclidean")
mask = np.loadtxt(scenario_folder + "/../mask.tab", delimiter=",")
name = scenario_folder.split(os.sep)[-2] + "_" + scenario_folder.split(
os.sep)[-1]
files = glob.glob(scenario_folder + "/*CH*rpt")
features = np.zeros(64)
for file in files:
# print(file)
CH = file.split(os.sep)[-1].split("-")[-1].split(".")[0][2:]
d = file.split(os.sep)[-1].split("-")[0].split(".")[0][1:]
# print(CH, d)
features[int(CH) - 1] = int(d)
print("Using distortion at scenario", glob.glob(scenario_folder))
print("Using distortion at scenario",
glob.glob(scenario_folder + "/D64*rpt")[-1])
distortions_data = open(glob.glob(scenario_folder + "/D64*rpt")[0],
"r").read()
distortions_coord_x = np.array(distortions_data[re.search(
"-------------------------------------------------------------------------------------------------",
distortions_data).end():re.search("Minimum", distortions_data).start(
)].split()[3:][::6]).astype(float)
distortions_coord_y = np.array(distortions_data[re.search(
"-------------------------------------------------------------------------------------------------",
distortions_data).end():re.search("Minimum", distortions_data).start(
)].split()[4:][::6]).astype(float)
distortions_coord_z = np.array(distortions_data[re.search(
"-------------------------------------------------------------------------------------------------",
distortions_data).end():re.search("Minimum", distortions_data).start(
)].split()[5:][::6]).astype(float)
zi = griddata(
(ini_coord_x + distortions_coord_x, ini_coord_z + distortions_coord_z),
ini_coord_y + distortions_coord_y, (xi, yi),
method='cubic')
distortions_coord_y = zi.flatten()
distortion_chuck_x = xi.flatten().reshape(xi.flatten().shape[0], 1)
distortion_chuck_y = distortions_coord_y.reshape(
distortions_coord_y.shape[0], 1)
distortion_chuck_z = yi.flatten().reshape(yi.flatten().shape[0], 1)
features_matrix = np.hstack((features.reshape(1, len(features)),
mask.reshape(1,
mask.shape[0] * mask.shape[1])))
features_matrix_dist = distances_matrix[features.astype(int) - 1].T
features_matrix_dist_corner = distances_matrix_corner[features.astype(int)
- 1].T
return [
np.hstack((features_matrix, features_matrix_dist,
features_matrix_dist_corner)),
(distortion_chuck_x, distortion_chuck_y, distortion_chuck_z),
[ini_coord_x, ini_coord_y, ini_coord_z], name
]
def pattern_sequence_masked_parallel_(
pattern_folder="C:/Users/mllamosa/Dropbox/2018/Applus/welding patterns all/"
):
folder_parent = pattern_folder
features_matrix = np.zeros(64)
features_matrix_dist = np.zeros(64)
pattern_names = []
file = pattern_folder + "Initial-Bottom.rpt"
[ini_coord_x, ini_coord_y, ini_coord_z] = coord_nodes(
pattern_folder=
"C:/Users/mllamosa/Dropbox/2018/Applus/welding patterns all/",
file_name="Initial-Bottom.rpt")
xi, yi = np.meshgrid(np.unique(ini_coord_x), np.unique(ini_coord_z))
centroids_matrix = np.mgrid[(max(ini_coord_x) - min(ini_coord_x)) /
16:max(ini_coord_x) -
(max(ini_coord_x) - min(ini_coord_x)) / 16:8j,
(max(ini_coord_x) - min(ini_coord_x)) /
16:max(ini_coord_x) -
(max(ini_coord_x) - min(ini_coord_x)) / 16:8j]
centroids_matrix = centroids_matrix.reshape(
2, centroids_matrix.shape[1] * centroids_matrix.shape[2]).T
coord_nodes_chunck = np.array([ini_coord_x, ini_coord_y, ini_coord_z]).T
distances_matrix = distance.cdist(
centroids_matrix,
coord_nodes_chunck.mean(axis=0)[[0, 2]].reshape(1, 2), "euclidean")
distances_matrix_corner = distance.cdist(
centroids_matrix, np.array([[min(ini_coord_x),
min(ini_coord_x)]]), "euclidean")
ll = 0
for folder_mask in glob.glob(folder_parent + "/Mask*"):
mask = np.loadtxt(folder_mask + "/mask.tab", delimiter=",")
k = 0
for folder in glob.glob(folder_mask + "/Scenario*"):
print(folder)
pattern_names.append(
folder_mask.split(os.sep)[-1] + "_" + folder.split(os.sep)[-1])
files = glob.glob(folder + "/*CH*rpt")
if len(files) > 0:
features = np.zeros(64)
for file in files:
# print(file)
CH = file.split(
os.sep)[-1].split("-")[-1].split(".")[0][2:]
d = file.split(os.sep)[-1].split("-")[0].split(".")[0][1:]
# print(CH, d)
features[int(CH) - 1] = int(d)
print("Using distortion at scenario",
glob.glob(folder + "/D64*"))
distortions_data = open(
glob.glob(folder + "/D64*rpt")[0], "r").read()
distortions_coord_x = np.array(distortions_data[re.search(
"-------------------------------------------------------------------------------------------------",
distortions_data).end(
):re.search("Minimum", distortions_data).start()].split()
[3:][::6]).astype(float)
distortions_coord_y = np.array(distortions_data[re.search(
"-------------------------------------------------------------------------------------------------",
distortions_data).end(
):re.search("Minimum", distortions_data).start()].split()
[4:][::6]).astype(float)
distortions_coord_z = np.array(distortions_data[re.search(
"-------------------------------------------------------------------------------------------------",
distortions_data).end(
):re.search("Minimum", distortions_data).start()].split()
[5:][::6]).astype(float)
zi = griddata((ini_coord_x + distortions_coord_x,
ini_coord_z + distortions_coord_z),
ini_coord_y + distortions_coord_y, (xi, yi),
method='cubic')
distortions_coord_y = zi.flatten()
distortion_chuck_x = xi.flatten().reshape(
xi.flatten().shape[0], 1)
distortion_chuck_y = distortions_coord_y.reshape(
distortions_coord_y.shape[0], 1)
distortion_chuck_z = yi.flatten().reshape(
yi.flatten().shape[0], 1)
if not k:
# print(features.reshape(len(features), 1).shape, mask.reshape(1, mask.shape[0]*mask.shape[1]).shape)
features_matrix = np.hstack(
(features.reshape(1, len(features)),
mask.reshape(1, mask.shape[0] * mask.shape[1])))
features_matrix_dist = distances_matrix[
features.astype(int) - 1].T
features_matrix_dist_corner = distances_matrix_corner[
features.astype(int) - 1].T
distortion_chuck_matrix_x = distortion_chuck_x
distortion_chuck_matrix_y = distortion_chuck_y
distortion_chuck_matrix_z = distortion_chuck_z
k = 1
else:
# print(features_matrix.shape, features.shape, mask.shape)
features_matrix = np.vstack(
(features_matrix,
np.hstack(
(features.reshape(1, len(features)),
mask.reshape(1,
mask.shape[0] * mask.shape[1])))))
features_matrix_dist = np.vstack(
(features_matrix_dist,
distances_matrix[features.astype(int) - 1].T))
features_matrix_dist_corner = np.vstack(
(features_matrix_dist_corner,
distances_matrix_corner[features.astype(int) - 1].T))
distortion_chuck_matrix_x = np.hstack(
(distortion_chuck_matrix_x, distortion_chuck_x))
distortion_chuck_matrix_y = np.hstack(
(distortion_chuck_matrix_y, distortion_chuck_y))
distortion_chuck_matrix_z = np.hstack(
(distortion_chuck_matrix_z, distortion_chuck_z))
else:
print("No data found in folder", folder)
if not ll:
features_matrix_full = np.hstack(
(features_matrix, features_matrix_dist,
features_matrix_dist_corner))
distortion_chuck_matrix_x_full = distortion_chuck_matrix_x
distortion_chuck_matrix_y_full = distortion_chuck_matrix_y
distortion_chuck_matrix_z_full = distortion_chuck_matrix_z
ll = 1
else:
features_matrix_full = np.vstack(
(features_matrix_full,
np.hstack((features_matrix, features_matrix_dist,
features_matrix_dist_corner))))
distortion_chuck_matrix_x_full = np.hstack(
(distortion_chuck_matrix_x_full, distortion_chuck_matrix_x))
distortion_chuck_matrix_y_full = np.hstack(
(distortion_chuck_matrix_y_full, distortion_chuck_matrix_y))
distortion_chuck_matrix_z_full = np.hstack(
(distortion_chuck_matrix_z_full, distortion_chuck_matrix_z))
print(features_matrix_full.shape)
# print(features_matrix.shape)
print(features_matrix_full.shape)
return features_matrix_full, [distortion_chuck_matrix_x_full, distortion_chuck_matrix_y_full,
distortion_chuck_matrix_z_full], \
[ini_coord_x, ini_coord_y, ini_coord_z], np.array(pattern_names)
def pattern_sequence(pattern_folder):
features_matrix = np.zeros(64)
features_matrix_dist = np.zeros(64)
pattern_names = []
[ini_coord_x, ini_coord_y,
ini_coord_z] = coord_nodes(pattern_folder=None,
file_name="Initial-Bottom.rpt")
xi, yi = np.meshgrid(np.unique(ini_coord_x), np.unique(ini_coord_z))
centroids_matrix = np.mgrid[(max(ini_coord_x) - min(ini_coord_x)) /
16:max(ini_coord_x) -
(max(ini_coord_x) - min(ini_coord_x)) / 16:8j,
(max(ini_coord_x) - min(ini_coord_x)) /
16:max(ini_coord_x) -
(max(ini_coord_x) - min(ini_coord_x)) / 16:8j]
centroids_matrix = centroids_matrix.reshape(
2, centroids_matrix.shape[1] * centroids_matrix.shape[2]).T
coord_nodes_chunck = np.array([ini_coord_x, ini_coord_y, ini_coord_z]).T
distances_matrix = distance.cdist(
centroids_matrix,
coord_nodes_chunck.mean(axis=0)[[0, 2]].reshape(1, 2), "euclidean")
distances_matrix_corner = distance.cdist(
centroids_matrix, np.array([[min(ini_coord_x),
min(ini_coord_x)]]), "euclidean")
k = 0
for folder in glob.glob(pattern_folder + "/Mask_0/Sce*"):
print(folder)
pattern_names.append(folder.split(os.sep)[-1])
files = glob.glob(folder + "/*CH*rpt")
if len(files) > 0:
features = np.zeros(64)
for file in files:
CH = file.split(os.sep)[-1].split("-")[-1].split(".")[0][2:]
d = file.split(os.sep)[-1].split("-")[0].split(".")[0][1:]
features[int(CH) - 1] = int(d)
print("Using distortion at scenario", glob.glob(folder + "/D64*"))
distortions_data = open(glob.glob(folder + "/D64*rpt")[0],
"r").read()
distortions_coord_x = np.array(distortions_data[re.search(
"-------------------------------------------------------------------------------------------------",
distortions_data).end():re.search("Minimum", distortions_data
).start()].split()[3:]
[::6]).astype(float)
distortions_coord_y = np.array(distortions_data[re.search(
"-------------------------------------------------------------------------------------------------",
distortions_data).end():re.search("Minimum", distortions_data
).start()].split()[4:]
[::6]).astype(float)
distortions_coord_z = np.array(distortions_data[re.search(
"-------------------------------------------------------------------------------------------------",
distortions_data).end():re.search("Minimum", distortions_data
).start()].split()[5:]
[::6]).astype(float)
zi = griddata((ini_coord_x + distortions_coord_x,
ini_coord_z + distortions_coord_z),
ini_coord_y + distortions_coord_y, (xi, yi),
method='cubic')
distortions_coord_y = zi.flatten()
distortion_chuck_x = xi.flatten().reshape(xi.flatten().shape[0], 1)
distortion_chuck_y = distortions_coord_y.reshape(
distortions_coord_y.shape[0], 1)
distortion_chuck_z = yi.flatten().reshape(yi.flatten().shape[0], 1)
if not k:
features_matrix = features
features_matrix_dist = distances_matrix[features.astype(int) -
1].T
features_matrix_dist_corner = distances_matrix_corner[
features.astype(int) - 1].T
distortion_chuck_matrix_x = distortion_chuck_x
distortion_chuck_matrix_y = distortion_chuck_y
distortion_chuck_matrix_z = distortion_chuck_z
k = 1
else:
features_matrix = np.vstack((features_matrix, features))
features_matrix_dist = np.vstack(
(features_matrix_dist,
distances_matrix[features.astype(int) - 1].T))
features_matrix_dist_corner = np.vstack(
(features_matrix_dist_corner,
distances_matrix_corner[features.astype(int) - 1].T))
distortion_chuck_matrix_x = np.hstack(
(distortion_chuck_matrix_x, distortion_chuck_x))
distortion_chuck_matrix_y = np.hstack(
(distortion_chuck_matrix_y, distortion_chuck_y))
distortion_chuck_matrix_z = np.hstack(
(distortion_chuck_matrix_z, distortion_chuck_z))
else:
print("No data found in folder", folder)
print(features_matrix.shape, features_matrix_dist.shape,
features_matrix_dist_corner.shape)
features_matrix = np.hstack(
(features_matrix, features_matrix_dist, features_matrix_dist_corner))
return features_matrix, [distortion_chuck_matrix_x, distortion_chuck_matrix_y, distortion_chuck_matrix_z], \
[ini_coord_x, ini_coord_y, ini_coord_z], np.array(pattern_names)