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
187
188
189
190
191
192
193
194
195
|
/*Copyright (C) 2015 Olivier Delaneau, Halit Ongen, Emmanouil T. Dermitzakis
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. If not, see <http://www.gnu.org/licenses/>.*/
#ifndef _RESIDUALIZER_H
#define _RESIDUALIZER_H
#define R_QR_TOLERANCE 1e-7
//ERROR CODE FOR COVARIATE PARSING
#define COV_OKAY 0
#define COV_MIXD 1
#define COV_DROP 2
#define COV_NVAR 3
#define COV_NCOV 4
#define COV_CORR 5
//STL INCLUDES
#include <vector>
#include <set>
#include <sstream>
#include <string>
//EIGEN INCLUDES
#include <Eigen/Dense>
#include <Eigen/LU>
//EIGEN NAMESPACE
using namespace Eigen;
class residualizer {
public:
unsigned int n_samples;
unsigned int n_covariates;
MatrixXd covarM;
MatrixXd PQR_Q;
MatrixXd PQR_Q_A;
ColPivHouseholderQR < MatrixXd > PQR;
residualizer(int _n_samples) : n_samples (_n_samples) {
covarM.resize(n_samples,1);
covarM.col(0) = VectorXd::Ones(n_samples);
n_covariates = 0;
}
~residualizer() {
n_samples = 0;
n_covariates = 0;
covarM.resize(0,0);
PQR_Q.resize(0,0);
PQR_Q_A.resize(0,0);
}
unsigned int push(vector < string > & covariate) {
set < string > factors;
set < unsigned int > i_yesmissing, i_nonmissing;
//MAP MISSING
for (int i = 0 ; i < covariate.size() ; i++) {
if (covariate[i] == "NA") i_yesmissing.insert(i);
else i_nonmissing.insert(i);
}
//TEST FOR NUMERIC
bool isNumeric = false, isAlphabetic = false;
for (set < unsigned int >::iterator itNM = i_nonmissing.begin(); itNM != i_nonmissing.end() ; ++itNM) {
float value;
std::istringstream in(covariate[*itNM]);
if (!(in >> value)) {
factors.insert(covariate[*itNM]);
isAlphabetic = true;
} else isNumeric = true;
}
if (isNumeric && isAlphabetic) return COV_MIXD;
//FILL IN VALUES
vector < vector < float > > additional_hcov;
if (factors.size() == 0) {
additional_hcov = vector < vector < float > > (1, vector < float > (n_samples, 0.0));
for (int i = 0 ; i < covariate.size() ; i++) additional_hcov[0][i] = std::stof(covariate[i]);
} else if (factors.size() > 1) {
factors.erase(factors.begin());
for (set < string > ::iterator itF = factors.begin(); itF != factors.end() ; itF++) {
additional_hcov.push_back(vector < float > (n_samples, 0.0));
for (int i = 0 ; i < n_samples ; i++) additional_hcov.back()[i] = (covariate[i] == (*itF));
}
} else return COV_DROP;
//IMPUTE MISSING
if (i_yesmissing.size() > 0) {
for (unsigned int c = 0 ; c < additional_hcov.size() ; c ++) {
double sum_row = 0;
for (set < unsigned int >::iterator itNM = i_nonmissing.begin(); itNM != i_nonmissing.end() ; ++itNM) sum_row += additional_hcov[c][*itNM];
for (set < unsigned int >::iterator itYM = i_yesmissing.begin(); itYM != i_yesmissing.end() ; ++itYM) additional_hcov[c][*itYM] = sum_row / i_nonmissing.size();
}
}
//ADD RESULTING COVARIATES
for (int c = 0 ; c < additional_hcov.size() ; c ++) if (!push(additional_hcov[c])) return COV_NVAR;
return COV_OKAY;
}
bool push(vector < float > & covariate) {
bool isVariable = false;
for (unsigned int e = 1 ; e < covariate.size() ; e ++) if (covariate[e] != covariate[e-1]) isVariable = true;
if (!isVariable) return false;
n_covariates ++;
covarM.conservativeResize(n_samples, n_covariates+1);
for(int i = 0 ; i < n_samples ; i ++) covarM(i, n_covariates) = covariate[i];
return true;
}
unsigned int build() {
if (n_covariates == 0) return COV_NCOV;
PQR = ColPivHouseholderQR<MatrixXd>(covarM);
PQR.setThreshold(R_QR_TOLERANCE);
if (PQR.rank() != n_covariates + 1) {
PQR_Q = PQR.householderQ();
PQR_Q_A = PQR.householderQ().adjoint();
return COV_CORR;
}
return COV_OKAY;
}
unsigned int residualize(vector < float > & data) {
if (n_covariates == 0) return COV_NCOV;
bool isVariable = false;
for (unsigned int e = 1 ; e < data.size() ; e ++) if (data[e] != data[e-1]) isVariable = true;
if (!isVariable) return COV_NVAR;
//FILL IN DATA
VectorXd counts(n_samples);
for(int i = 0; i < n_samples ; i ++) counts(i) = data[i];
//CORRECTION
if (PQR.rank() == n_covariates + 1) {
VectorXd m_coef = PQR.solve(counts);
VectorXd fitted = covarM * m_coef;
VectorXd e = counts - fitted;
for (int i = 0; i < e.size(); i ++) data[i] = e(i);
} else {
VectorXd effects(PQR_Q_A * counts);
effects.tail(n_samples - PQR.rank()).setZero();
VectorXd fitted = PQR_Q * effects;
VectorXd e = counts - fitted;
for (int i = 0; i < e.size(); i ++) data[i] = (float)e(i);
}
return COV_OKAY;
}
unsigned int residualize(float * data) {
if (n_covariates == 0) return COV_NCOV;
bool isVariable = false;
for (unsigned int e = 1 ; e < n_samples ; e ++) if (data[e] != data[e-1]) isVariable = true;
if (!isVariable) return COV_NVAR;
//FILL IN DATA
VectorXd counts(n_samples);
for(int i = 0; i < n_samples ; i ++) counts(i) = data[i];
//CORRECTION
if (PQR.rank() == n_covariates + 1) {
VectorXd m_coef = PQR.solve(counts);
VectorXd fitted = covarM * m_coef;
VectorXd e = counts - fitted;
for (int i = 0; i < e.size(); i ++) data[i] = e(i);
} else {
VectorXd effects(PQR_Q_A * counts);
effects.tail(n_samples - PQR.rank()).setZero();
VectorXd fitted = PQR_Q * effects;
VectorXd e = counts - fitted;
for (int i = 0; i < e.size(); i ++) data[i] = (float)e(i);
}
return COV_OKAY;
}
};
#endif
|
