-
Notifications
You must be signed in to change notification settings - Fork 2
/
functions.py
186 lines (159 loc) · 8.26 KB
/
functions.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
"""
SPECTRA PROCESSING
Copyright (C) 2020 Josef Brandt, University of Gothenborg.
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program, see COPYING.
If not, see <https://www.gnu.org/licenses/>.
"""
from typing import List, Tuple, Dict, Set
import numpy as np
from sklearn.decomposition import PCA
from sklearn.cluster import k_means
from sklearn.preprocessing import StandardScaler
from sklearn.metrics import classification_report
import importData as io
def compareResultLists(trueList: List[str], estimatedList: List[str]) -> Tuple[float, float, float, float]:
"""
Compares two assignment lists.
:param trueList: List of true assignments
:param estimatedList: List of estimated assignments
:returns: Tuple(agreement in percent, Dictionary of agreements in percent per type of assignment)
"""
assert len(trueList) == len(estimatedList)
# Do some replacements to facilitate evaluation
replaceDict: dict = {'ldpe': 'pe',
'hdpe': 'pe'}
plastError = getPlasticError(trueList, estimatedList)
for i in range(len(trueList)):
trueList[i] = trueList[i].lower()
estimatedList[i] = estimatedList[i].lower()
if trueList[i] in replaceDict.keys():
trueList[i] = replaceDict[trueList[i]]
if estimatedList[i] in replaceDict.keys():
estimatedList[i] = replaceDict[estimatedList[i]]
# remove unknowns (we are not interested in statistics about them)
cleanTrueList, cleanEstimateList = [], []
for i, entry in enumerate(trueList):
if entry != "unknown":
cleanTrueList.append(entry)
cleanEstimateList.append(estimatedList[i])
trueList = cleanTrueList
estimatedList = cleanEstimateList
uniques: List[str] = list(np.unique(trueList))
report: Dict[str, Dict] = classification_report(trueList, estimatedList, output_dict=True, zero_division=0)
precisionDict: Dict[str, float] = {}
recallDict: Dict[str, float] = {}
for cls, results in report.items():
if cls in uniques:
precisionDict[cls] = round(results["precision"], 2)
recallDict[cls] = round(results["recall"], 2)
avgPrecision = np.mean(list(precisionDict.values())) * 100
avgRecall = np.mean(list(recallDict.values())) * 100
if (avgRecall + avgPrecision) == 0:
avgF1 = 0
else:
avgF1 = 2*(avgPrecision*avgRecall) / (avgPrecision + avgRecall)
return avgPrecision, avgRecall, avgF1, plastError
def getPlasticError(trueList: List[str], estimatedList: List[str]) -> float:
nonPlastNames: Set[str] = set(io.getNonPlasticNames() + ["unknown"])
from collections import Counter
# print("True unknowns:", Counter(trueList).get("unknown"), "Estimated unknowns:", Counter(estimatedList).get("unknown"))
plastCountTrue, plastCountEstimated = 0, 0
for true, estim in zip(trueList, estimatedList):
if true not in nonPlastNames:
plastCountTrue += 1
if estim not in nonPlastNames:
plastCountEstimated += 1
error: float = round((plastCountEstimated - plastCountTrue) / plastCountTrue * 100)
# print("TruePlastCount:", plastCountTrue, "Estimated PlastCount", plastCountEstimated, "PlastError:", error)
return error
def getNMostDifferentSpectra(assignments: List[str], spectra: np.ndarray, n: int) -> Tuple[List[str], np.ndarray]:
"""
Takes a set of spectra and returns the indices of the n spectra that are furthest apart from each other,
in terms of the first two PCA components.
:param assignments: List of M-1 spectra
:param spectra: (NxM) array of spec of M-1 spectra with N wavenumbers (wavenumbers in first column).
:param n: Desired number of spectra to keep
:return: Tuple[shortened AssignmentList, shortened Spectra array]
"""
maxIndex = np.argmin(np.abs(spectra[:, 0] - 2000)) # only go up to 2000 cm-1, above it's so unspecific...
intensities = spectra[:maxIndex, 1:]
intensities = StandardScaler().fit_transform(intensities.transpose())
indices: List[int] = []
pca: PCA = PCA(n_components=2, random_state=42)
princComps: np.ndarray = pca.fit_transform(intensities)
centers = k_means(princComps, n, random_state=42)[0]
for i in range(n):
distances = np.linalg.norm(princComps-centers[i, :], axis=1)
indices.append(int(np.argmin(distances)))
# FOR DEBUG DISPLAY
# import matplotlib.pyplot as plt
# plt.scatter(princComps[:, 0], princComps[:, 1], color='lightgray')
# plt.scatter(princComps[indices, 0], princComps[indices, 1], color='black')
# plt.xlabel('PC 1')
# plt.ylabel('PC 2')
# plt.title(f'Chosing {n} out of {princComps.shape[0]} spectra')
# plt.show(block=True)
assignments = [assignments[i] for i in indices]
indices = [0] + [i + 1 for i in indices]
spectra = spectra[:, indices]
assert len(assignments) == spectra.shape[1] - 1
return assignments, spectra
def mapSpectrasetsToSameWavenumbers(set1: np.ndarray, set2: np.ndarray) -> Tuple[np.ndarray, np.ndarray]:
"""
Matches two sets of spectra to each other, so that both have the same wavenumber axis (the shorter set is used)
:param set1: np.ndarray shape(N, M) with N wavenumbers and M-1 spectra, wavenumbers in first column
:param set2: np.ndarray shape(K, L) with K wavenumbers and L-1 spectra, wavenumbers in first column
:return: Tuple of set1 and set2, with same wavenumber axis.
"""
shorterWavenums, longerWavenums = set1[:, 0], set2[:, 0]
set1IsShorter: bool = True
if len(shorterWavenums) > len(longerWavenums):
shorterWavenums, longerWavenums = longerWavenums, shorterWavenums
set1IsShorter = False
newSet1: np.ndarray = set1 if set1IsShorter else set2
newSet2: np.ndarray = np.zeros((set1.shape[0], set2.shape[1])) if set1IsShorter else np.zeros((set2.shape[0], set1.shape[1]))
newSet2[:, 0] = shorterWavenums
for i, num in enumerate(shorterWavenums):
correspondInd = np.argmin(np.abs(longerWavenums - num))
newSet2[i, :] = set2[correspondInd, :] if set1IsShorter else set1[correspondInd, :]
returnTuple: Tuple[np.ndarray, np.ndarray] = (newSet1, newSet2) if set1IsShorter else (newSet2, newSet1)
assert returnTuple[0].shape[1] == set1.shape[1] and returnTuple[1].shape[1] == set2.shape[1]
assert returnTuple[0].shape[0] == returnTuple[1].shape[0] == len(shorterWavenums)
return returnTuple
def remapSpecArrayToWavenumbers(spectra: np.ndarray, wavenumbers: np.ndarray) -> np.ndarray:
"""
Takes a spectrum array and maps it to the currently present spectra.
:param spectra: (N, M) shape array of M-1 spectra with wavenumbs in first column
:param wavenumbers: The wavenumbers to map to. If None, the wavenumbers of the currently present spectra set
is used.
:return: shape (L, M) shape spectrum with new wavenumber axis
"""
newSpecs: np.ndarray = np.zeros((len(wavenumbers), spectra.shape[1]))
newSpecs[:, 0] = wavenumbers
for i in range(spectra.shape[1]-1):
newSpecs[:, i+1] = remapSpectrumToWavenumbers(spectra[:, [0, i+1]], wavenumbers)[:, 1]
return newSpecs
def remapSpectrumToWavenumbers(spectrum: np.ndarray, wavenumbers: np.ndarray = None) -> np.ndarray:
"""
Takes a spectrum array and maps it to the currently present spectra.
:param spectrum: (N, 2) shape spectrum with wavenumbs in first column
:param wavenumbers: The wavenumbers to map to. If None, the wavenumbers of the currently present spectra set
is used.
:return: shape (M, 2) shape spectrum with new wavenumber axis
"""
newSpec = np.zeros((len(wavenumbers), 2))
newSpec[:, 0] = wavenumbers
for i in range(len(wavenumbers)):
clostestIndex = np.argmin(np.abs(spectrum[:, 0] - wavenumbers[i]))
newSpec[i, 1] = spectrum[clostestIndex, 1]
return newSpec