new output eigvec2, which is plink2-like

CHANGELOG.org

@@ -1,3 +1,7 @@
+- v0.4.3
+  - new output =eigvecs2= for plink input
+  - fix ld clumping and sort snps by pvalues
+
 - v0.4.2
   - fix slowness due to the ld logics
   - add ld-clump functions

src/Main.cpp

@@ -109,6 +109,9 @@ int main(int argc, char * argv[])
     cao << tick.date() << "total elapsed wall time: " << tick.abstime() << " seconds" << endl;
     cao << tick.date() << "eigenvecs and eigenvals are saved. have a nice day. bye!\n";
 
+    if(params.file_t == FileType::PLINK)
+        make_plink2_eigenvec_file(params.k, params.fileout + ".eigvecs2", params.fileout + ".eigvecs",
+                                  params.filein + ".fam");
     delete data;
 
     return 0;

src/Utils.cpp

@@ -278,3 +278,19 @@ void get_snp_pos_bim(const std::string & filebim,
     chr_pos_end.push_back(i - 1); // add the last SNP
 }
 
+void make_plink2_eigenvec_file(int K, std::string fout, const std::string & fin, const std::string & fam)
+{
+    std::ifstream ifam(fam);
+    std::ifstream ifin(fin);
+    std::ofstream ofs(fout);
+    ofs << "#FID\tIID";
+    for(int i = 0; i < K; i++) ofs << "\tPC" << i + 1;
+    ofs << "\n";
+    std::string line1, line2, sep{" \t"};
+    while(getline(ifam, line1))
+    {
+        auto tokens = split_string(line1, sep);
+        getline(ifin, line2);
+        ofs << tokens[0] + "\t" + tokens[1] + "\t" << line2 << std::endl;
+    }
+}

src/Utils.hpp

@@ -56,7 +56,6 @@ inline std::vector<size_t> sortidx(const std::vector<T> & v)
     return idx;
 }
 
-
 struct Line
 {
     std::string data;
@@ -107,4 +106,6 @@ void get_snp_pos_bim(const std::string & filebim,
                      bool header = false,
                      Int1D idx = Int1D{0, 3});
 
+void make_plink2_eigenvec_file(int K, std::string fout, const std::string & fin, const std::string & fam);
+
 #endif // PCAONE_UTILES_