diff --git a/src/common.py b/src/common.py
index 5720c506b695ef4928302d66c2bdf4205e56ae5e..92d27f25c22f702ce991850b649e690bcf29e552 100644
--- a/src/common.py
+++ b/src/common.py
@@ -18,9 +18,12 @@ image_path = Path('./images/img-sky.jpg')
import os
from shutil import rmtree
-from pandas import DataFrame
+from pandas import DataFrame, concat
from PIL import Image
import plotly.express as px
+from tempfile import NamedTemporaryFile
+import numpy as np
+from datetime import datetime
from utils.data_parsing import JcampParser, CsvParser
diff --git a/src/pages/1-samples_selection.py b/src/pages/1-samples_selection.py
index 4ec60a16da8f49c52356cafefde96e7826c9c6f2..3314d99a0c20720b5e245e18b8cc15ff50d0dc04 100644
--- a/src/pages/1-samples_selection.py
+++ b/src/pages/1-samples_selection.py
@@ -312,6 +312,7 @@ elif labels:
# Strategy 0
case 'center':
# list samples at clusters centers - Use sklearn.metrics.pairwise_distances_argmin if you want more than 1 sample per cluster
+ from sklearn.metrics import pairwise_distances_argmin_min
closest, _ = pairwise_distances_argmin_min(clu_centers, new_tcr)
selected_samples_idx = np.array(new_tcr.index)[list(closest)]
selected_samples_idx = selected_samples_idx.tolist()
@@ -393,6 +394,7 @@ if not t.empty:
fig_export = {}
# export 2D scores plot
if len(axis)== 3:
+ from itertools import combinations
comb = [i for i in combinations(np.arange(len(axis)), 2)]
subcap = ['a','b','c']
for i in range(len(comb)):
@@ -456,7 +458,8 @@ if not spectra.empty:
xp = np.dot(t,p.T)
# Q residuals: Q residuals represent the magnitude of the variation remaining in each sample after projection through the model
residuals = np.diag(np.subtract(xc.to_numpy(), xp)@ np.subtract(xc.to_numpy(), xp).T)
- tresh4 = sc.stats.chi2.ppf(0.05, df = len(axis))
+ from scipy.stats import chi2
+ tresh4 = chi2.ppf(0.05, df = len(axis))
# color with metadata
if colfilter:
@@ -514,10 +517,11 @@ if not spectra.empty:
hotelling = t.var(axis = 1)
# Q residuals: Q residuals represent the magnitude of the variation remaining in each sample after projection through the model
residuals = np.diag(np.subtract(xc.to_numpy(), xp)@ np.subtract(xc.to_numpy(), xp).T)
-
- fcri = sc.stats.f.isf(0.05, 3, n_samples)
+
+ from scipy.stats import f, chi2
+ fcri = f.isf(0.05, 3, n_samples)
tresh0 = (3 * (n_samples ** 2 - 1) * fcri) / (n_samples * (n_samples - 3))
- tresh1 = sc.stats.chi2.ppf(0.05, df = 3)
+ tresh1 = chi2.ppf(0.05, df = 3)
hotelling_plot = px.scatter(t, x = hotelling, y = residuals, color=labels if list(labels) else None,
color_discrete_sequence= custom_color_palette)
hotelling_plot.add_scatter(x = hotelling[selected_samples_idx] , y = residuals[selected_samples_idx],
diff --git a/src/utils/clustering.py b/src/utils/clustering.py
index 4f9a9c5a7aaf00d2eb989419187274c2a671d5d0..d367d042d3ce61fb78be16703dab7b24f41cfaaa 100644
--- a/src/utils/clustering.py
+++ b/src/utils/clustering.py
@@ -1,3 +1,8 @@
+import numpy as np
+from pandas import DataFrame
+from sklearn.cluster import KMeans
+
+
#~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ kmeans ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
class Sk_Kmeans:
diff --git a/src/utils/data_handling.py b/src/utils/data_handling.py
index 28d2029dfaae47af75e70c6b4bbe41c4d6bd6128..1c222742d66d71f3b931fe1279f4d6960eed0004 100644
--- a/src/utils/data_handling.py
+++ b/src/utils/data_handling.py
@@ -54,6 +54,8 @@ def list_files(mypath, import_type):
def standardize(X, center = True, scale = False):
+ from pandas import DataFrame
+ from sklearn.preprocessing import StandardScaler
sk = StandardScaler(with_mean=center, with_std = scale)
sc = DataFrame(sk.fit_transform(X), index = X.index, columns = X.columns)
return sc
diff --git a/src/utils/data_parsing.py b/src/utils/data_parsing.py
index d78ef2b0cffe787a64fd34a858e954382dc88c5b..9e536071725fd6f4647c34e80be24bfd6f0c4df4 100644
--- a/src/utils/data_parsing.py
+++ b/src/utils/data_parsing.py
@@ -13,6 +13,7 @@ class JcampParser:
self.__wl = self.__list_of_blocks[0]["x"] # Wavelengths/frequencies/range
def parse(self):
+ import numpy as np
# Start retreiving the data
specs = np.zeros((self.__nb, len(self.__list_of_blocks[0]["y"])), dtype=float) # preallocate a np matrix for sotoring spectra
self.idx = np.arange(self.__nb) # This list is designed to store samples name
@@ -43,6 +44,7 @@ class JcampParser:
}
self.__met[f'{i}'] = block_met
+ from pandas import DataFrame
self.metadata_ = DataFrame(self.__met).T
self.spectra = DataFrame(np.fliplr(specs), columns= self.__wl[::-1], index = self.metadata_['name']) # Storing spectra in a dataframe
@@ -55,6 +57,7 @@ class JcampParser:
n_elements = a.count('(')
## Get the name of analyzed chamical elements
+ import re
elements_name = []
for match in re.findall(self.pattern, a):
elements_name.append(match[0])
@@ -71,6 +74,8 @@ class JcampParser:
### Method for retrieving the concentration of a single sample
def conc(self,sample):
+ import re
+ import numpy as np
prep = '\n'.join(line for line in sample.split('\n') if "NCU" not in line and "<<undef>>" not in line)
c = []
for match in re.findall(self.pattern, prep):
diff --git a/src/utils/dim_reduction.py b/src/utils/dim_reduction.py
index b6a13bd8eead619710ecbeb05440ae1587ec85c4..0c112221aa1057e72407544241e96670ec7f8fcd 100644
--- a/src/utils/dim_reduction.py
+++ b/src/utils/dim_reduction.py
@@ -1,10 +1,12 @@
from utils.data_handling import *
-
+from pandas import DataFrame
+import numpy as np
# ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ pca ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~#
class LinearPCA:
def __init__(self, X, Ncomp=10):
+ from sklearn.decomposition import PCA
## input matrix
self.__x = np.array(X)
## set the number of components to compute and fit the model
@@ -67,11 +69,13 @@ class Umap:
self.categorical_data_encoded = cat_data
elif len(cat_data) > 0:
self.categorical_data = cat_data
+ from sklearn.preprocessing import LabelEncoder
self.le = LabelEncoder()
self.categorical_data_encoded = self.le.fit_transform(self.categorical_data)
else:
self.categorical_data_encoded = None
+ from umap import UMAP
self.model = UMAP(n_neighbors=20, n_components=3, min_dist=0.0, )#random_state=42,)
self.model.fit(self.numerical_data, y = self.categorical_data_encoded)
self.scores_raw = self.model.transform(self.numerical_data)
@@ -97,6 +101,7 @@ class Nmf:
self.__ncp = Ncomp
# Fit PCA model
+ from sklearn.decomposition import NMF
Mo = NMF(n_components=self.__ncp, init=None, solver='cd', beta_loss='frobenius',
tol=0.0001, max_iter=300, random_state=None, alpha_W=0.0, alpha_H='same',
l1_ratio=0.0, verbose=0, shuffle=False)
diff --git a/src/utils/visualize.py b/src/utils/visualize.py
index b560be60548a90cdf9a4a728ac1de63632b6a5de..5ac80c34c96ab0bd1df4308a4f9ee36581d50e37 100644
--- a/src/utils/visualize.py
+++ b/src/utils/visualize.py
@@ -26,6 +26,9 @@ def pred_hist(pred):
# ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ predictions histogram ~~~~~~~~~~~~~~~~~~~~~~~~~~
@st.cache_data
def plot_spectra(specdf, xunits, yunits):
+ import matplotlib.pyplot as plt
+ import numpy as np
+
fig, ax = plt.subplots(figsize = (30,7))
if isinstance(specdf.columns[0], str):
specdf.T.plot(legend=False, ax = ax, color = '#2474b4')