1 | // $Id: qQN.cc 1050 2009-05-06 09:11:59Z jari $ |
---|
2 | |
---|
3 | /* |
---|
4 | Copyright (C) 2009 Jari Häkkinen |
---|
5 | |
---|
6 | This file is part of the Normalizers plug-in package for BASE |
---|
7 | (net.sf.based.normalizers). The package is available at |
---|
8 | http://baseplugins.thep.lu.se/ BASE main site is |
---|
9 | http://base.thep.lu.se/ |
---|
10 | |
---|
11 | This is free software; you can redistribute it and/or modify it |
---|
12 | under the terms of the GNU General Public License as published by |
---|
13 | the Free Software Foundation; either version 3 of the License, or |
---|
14 | (at your option) any later version. |
---|
15 | |
---|
16 | The software is distributed in the hope that it will be useful, but |
---|
17 | WITHOUT ANY WARRANTY; without even the implied warranty of |
---|
18 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
---|
19 | General Public License for more details. |
---|
20 | |
---|
21 | You should have received a copy of the GNU General Public License |
---|
22 | along with this program. If not, see <http://www.gnu.org/licenses/>. |
---|
23 | */ |
---|
24 | |
---|
25 | #include <config.h> // this header file is created by configure |
---|
26 | |
---|
27 | #include <yat/normalizer/ColumnNormalizer.h> |
---|
28 | #include <yat/normalizer/qQuantileNormalizer.h> |
---|
29 | |
---|
30 | #include <yat/utility/CommandLine.h> |
---|
31 | #include <yat/utility/MatrixWeighted.h> |
---|
32 | #include <yat/utility/OptionHelp.h> |
---|
33 | #include <yat/utility/OptionInFile.h> |
---|
34 | #include <yat/utility/OptionOutFile.h> |
---|
35 | #include <yat/utility/OptionSwitch.h> |
---|
36 | |
---|
37 | #include <cstdlib> |
---|
38 | #include <fstream> |
---|
39 | #include <iostream> |
---|
40 | #include <stdexcept> |
---|
41 | |
---|
42 | using namespace theplu::yat::normalizer; |
---|
43 | using namespace theplu::yat::utility; |
---|
44 | |
---|
45 | |
---|
46 | void create_target_and_fix_nans(std::vector<double>&, MatrixWeighted&); |
---|
47 | void create_target_and_fix_nans(std::vector<double>&, MatrixWeighted&, |
---|
48 | const std::string&); |
---|
49 | |
---|
50 | void exp_numbers(MatrixWeighted&); |
---|
51 | template<typename Iterator> void log_numbers(Iterator first, Iterator last); |
---|
52 | void log_numbers(MatrixWeighted&); |
---|
53 | |
---|
54 | /** |
---|
55 | writes the data values in the matrix ignoring the weights, i.e., |
---|
56 | produces the same output as the Matrix output operator does. |
---|
57 | */ |
---|
58 | std::ostream& operator<< (std::ostream&, const MatrixWeighted&); |
---|
59 | |
---|
60 | |
---|
61 | int main(int argc, char* argv[]) |
---|
62 | { |
---|
63 | CommandLine cmd; |
---|
64 | OptionInFile assay(cmd, "assay-data", "assay annotations"); |
---|
65 | OptionInFile indata(cmd, "in-data", "data to be normalized"); |
---|
66 | OptionOutFile outdata(cmd, "out-data", "normalized data"); |
---|
67 | OptionHelp help(cmd); |
---|
68 | help.synopsis()=(std::string("See ") + |
---|
69 | "http://baseplugins.thep.lu.se/net.sf.basedb.normalizers " + |
---|
70 | "for\ndetails on this program\n"); |
---|
71 | OptionSwitch version(cmd, "version", "output version and exit"); |
---|
72 | std::stringstream copyright; |
---|
73 | copyright << PACKAGE_STRING << '\n' |
---|
74 | << "Copyright (C) 2009 Jari Häkkinen\n\n" |
---|
75 | << "This is free software see the source for copying " |
---|
76 | << "conditions. There is NO\nwarranty; not even for " |
---|
77 | << "MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.\n"; |
---|
78 | try { |
---|
79 | cmd.parse(argc, argv); |
---|
80 | } |
---|
81 | catch (cmd_error& e) { |
---|
82 | if (version.present()) { |
---|
83 | std::cout << copyright.str(); |
---|
84 | return EXIT_SUCCESS; |
---|
85 | } |
---|
86 | std::cout << e.what() << std::endl; |
---|
87 | return EXIT_FAILURE; |
---|
88 | } |
---|
89 | if (version.present()) { |
---|
90 | std::cout << copyright.str(); |
---|
91 | return EXIT_SUCCESS; |
---|
92 | } |
---|
93 | std::ifstream* infile=NULL; |
---|
94 | std::streambuf* cin_buffer=NULL; |
---|
95 | if (indata.present()) { |
---|
96 | infile=new std::ifstream(indata.value().c_str()); |
---|
97 | cin_buffer = std::cin.rdbuf(); // save cin's input buffer |
---|
98 | std::cin.rdbuf(infile->rdbuf()); |
---|
99 | } |
---|
100 | MatrixWeighted m(std::cin,'\t'); |
---|
101 | if (indata.present()) { |
---|
102 | std::cin.rdbuf(cin_buffer); // restore old input buffer |
---|
103 | infile->close(); |
---|
104 | delete infile; |
---|
105 | } |
---|
106 | |
---|
107 | std::vector<double> target; |
---|
108 | ( assay.present() ? create_target_and_fix_nans(target,m,assay.value()) : |
---|
109 | create_target_and_fix_nans(target,m) ); |
---|
110 | log_numbers(target.begin(),target.end()); |
---|
111 | log_numbers(m); |
---|
112 | qQuantileNormalizer qqn(target.begin(), target.end(), 100); |
---|
113 | ColumnNormalizer<qQuantileNormalizer> cn(qqn); |
---|
114 | MatrixWeighted result(m.rows(),m.columns()); |
---|
115 | cn(m,result); |
---|
116 | exp_numbers(result); |
---|
117 | |
---|
118 | std::ofstream* outfile=NULL; |
---|
119 | std::streambuf* cout_buffer = std::cout.rdbuf(); |
---|
120 | if (outdata.present()) { |
---|
121 | outfile=new std::ofstream(outdata.value().c_str()); |
---|
122 | cout_buffer = std::cout.rdbuf(); // save cout's output buffer |
---|
123 | std::cout.rdbuf(outfile->rdbuf()); |
---|
124 | } |
---|
125 | std::cout << result << std::endl; |
---|
126 | if (outdata.present()) { |
---|
127 | std::cout.rdbuf(cout_buffer); // restore old output buffer |
---|
128 | outfile->close(); |
---|
129 | delete outfile; |
---|
130 | } |
---|
131 | |
---|
132 | return EXIT_SUCCESS; |
---|
133 | } |
---|
134 | |
---|
135 | |
---|
136 | void create_target_and_fix_nans(std::vector<double>& t, MatrixWeighted& m, |
---|
137 | const std::string& assay) |
---|
138 | { |
---|
139 | std::ifstream is(assay.c_str()); |
---|
140 | std::string line; |
---|
141 | size_t column=0; |
---|
142 | std::vector<size_t> column_contribs(m.rows(),0); |
---|
143 | std::vector<double> temp_target(m.rows(),0.0); |
---|
144 | while (getline(is, line)) { |
---|
145 | size_t found=line.find("yes"); |
---|
146 | |
---|
147 | if (found!=std::string::npos) { |
---|
148 | // use this assay as a part of reference |
---|
149 | for (size_t row=0; row<m.rows(); ++row) |
---|
150 | if (m(row,column).weight()) { // weight either 0 or 1 |
---|
151 | temp_target[row]+=m(row,column).data(); |
---|
152 | ++column_contribs[row]; |
---|
153 | } |
---|
154 | else |
---|
155 | m(row,column).data()=0; |
---|
156 | } |
---|
157 | else |
---|
158 | // this assay as not a part of the reference but fix nan |
---|
159 | for (size_t row=0; row<m.rows(); ++row) |
---|
160 | if (!m(row,column).weight()) // weight either 0 or 1 |
---|
161 | m(row,column).data()=0; |
---|
162 | |
---|
163 | ++column; |
---|
164 | if (column>m.columns()) |
---|
165 | throw std::runtime_error("Too many annotation columns wrt data matrix"); |
---|
166 | } |
---|
167 | t.reserve(m.rows()); |
---|
168 | for (size_t row=0; row<m.rows(); ++row) |
---|
169 | if (column_contribs[row]) |
---|
170 | t.push_back(temp_target[row]/=column_contribs[row]); |
---|
171 | if (!t.size()) |
---|
172 | throw std::runtime_error("Not a well defined reference, aborting"); |
---|
173 | } |
---|
174 | |
---|
175 | |
---|
176 | void create_target_and_fix_nans(std::vector<double>& t, MatrixWeighted& m) |
---|
177 | { |
---|
178 | std::vector<double> temp_target(m.rows(),0.0); |
---|
179 | t.reserve(m.rows()); |
---|
180 | for (size_t row=0; row<m.rows(); ++row) { |
---|
181 | size_t column_contribs=0; |
---|
182 | for (size_t column=0; column<m.columns(); ++column) |
---|
183 | if (m(row,column).weight()) { // weight either 0 or 1 |
---|
184 | temp_target[row]+=m(row,column).data(); |
---|
185 | ++column_contribs; |
---|
186 | } |
---|
187 | else |
---|
188 | m(row,column).data()=0; |
---|
189 | if (column_contribs) |
---|
190 | t.push_back(temp_target[row]/=column_contribs); |
---|
191 | } |
---|
192 | if (!t.size()) |
---|
193 | throw std::runtime_error("Not a well defined reference, aborting"); |
---|
194 | } |
---|
195 | |
---|
196 | |
---|
197 | void exp_numbers(MatrixWeighted& m) |
---|
198 | { |
---|
199 | MatrixWeighted::iterator i=m.begin(); |
---|
200 | while (i!=m.end()) { |
---|
201 | i->data() =std::exp(i->data()); |
---|
202 | ++i; |
---|
203 | } |
---|
204 | } |
---|
205 | |
---|
206 | |
---|
207 | template<typename Iterator> void log_numbers(Iterator i, Iterator last) |
---|
208 | { |
---|
209 | while (i!=last) { |
---|
210 | *i=std::log(*i); |
---|
211 | ++i; |
---|
212 | } |
---|
213 | } |
---|
214 | |
---|
215 | |
---|
216 | void log_numbers(MatrixWeighted& m) |
---|
217 | { |
---|
218 | MatrixWeighted::iterator i=m.begin(); |
---|
219 | while (i!=m.end()) { |
---|
220 | i->data() =std::log(i->data()); |
---|
221 | ++i; |
---|
222 | } |
---|
223 | } |
---|
224 | |
---|
225 | |
---|
226 | std::ostream& operator<< (std::ostream& s, const MatrixWeighted& m) |
---|
227 | { |
---|
228 | s.setf(std::ios::dec); |
---|
229 | s.precision(12); |
---|
230 | for(size_t i=0, j=0; i<m.rows(); i++) |
---|
231 | for (j=0; j<m.columns(); j++) { |
---|
232 | if (m(i,j).weight()) |
---|
233 | s << m(i,j).data(); |
---|
234 | else |
---|
235 | s << "nan"; |
---|
236 | if (j<m.columns()-1) |
---|
237 | s << s.fill(); |
---|
238 | else if (i<m.rows()-1) |
---|
239 | s << "\n"; |
---|
240 | } |
---|
241 | return s; |
---|
242 | } |
---|