mex matlab function c++ gives wrong answer relative to R c++ function

See also questions close to this topic

• Is volatile an illegal instruction in c++ on macos (related to building opencv)?

  I am trying to build OpenCV 4.1.0 on Mac by following this blog: https://gist.github.com/jruizvar/0535fb8612afb105e0eef64051dc0d00.

  Here is the command line for the preparatory cmake - which is successful:

  cmake -DOPENCV_EXTRA_MODULES_PATH=/git/opencv_contrib-4.1.0/modules -DPYTHON_DEFAULT_EXECUTABLE=python3 -DCMAKE_CXX_COMPILER=/usr/local/gcc-9.1/bin/g++-9.1 -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=/usr/local -DBUILD_opencv_python2=False -DBUILD_opencv_python3=True -DPYTHON3_EXECUTABLE=python3 ../
  

  Then the make

  make -j2
  

  That fails:

  /git/opencv-4.1.0/3rdparty/openexr/Half/half.cpp:71:5: internal compiler error: Illegal instruction: 4
     71 |     volatile float f = 1e10;
        |     ^~~~~~~~
  

  Tips from anyone who has successfully built this version of openCV on mac (and with opencv_contrib) would be especially appreciated.
• Running antlr4 parser for c++ on grammar file shows error 33: missing code generation template NonLocalAttrRefHeader

  I am relatively new to ANTLR, I have a current project that needs to be merged from ANTLR3 (version 3.5) to ANTLR4. I have gone thru the book and tried the demo, this all works fine, but my own project gives me the following problem:

  After converting a ANTRL3 project to ANTLR4 project (resolving all warnings and errors) I was able to build the lexer.h and lexer.cpp file but the following errors come up: error(33): missing code generation template NonLocalAttrRefHeader error(33): missing code generation template SetNonLocalAttrHeader (about 50 times). I haven't been able to find any references of these templates anywhere. Is there anybody who can shed any light on these error messages? Because they don't say anything about line no's or reference any other code I'm completely in in the dark where to look.

  I've set up a test environment, testing the demo g4 files. I have pulled the g4 file out of my (VS2017) project and tried it seperately using batch files.

  Because of the lack of references I can't show the actual piece of code that is the cause. I have tried a partial parse, but I haven't been able to get any clues from that.

  These errors are shown: error(33): missing code generation template NonLocalAttrRefHeader error(33): missing code generation template SetNonLocalAttrHeader
• Error for different type of list declaration in STL c++

  list<int>L[]=new list<int>[20];
  

  This gives an error " array initializer must be an initializer list" What is wrong with initializing list like:

  list<int> L[100];

  Also does list[] L mean anything

  list<int> *L=new list<int>[20] Why does this work ?

  Please give me brief explaination for each of the cases if possible as I am new to STL
• Why is .png file not looking clear after arranging using grid.arrange?

  I am trying to arrange 9 png files (which I had created using plot() function) in 2 columns. But these files are not looking clear after using grid.arrange to arrange them.

  I think if page size is increased, the images will be clearly visible. Or there might be some problem with the size of my plotes converted to .png.

  Please suggest me a solution.

  Following is the code I am trying:

  install.packages('png')
  install.packages('gridExtra')
  library(png)
  library(grid)
  library(gridExtra)
  
  p1 <- readPNG('p1.png')
  p2 <- readPNG('p2.png')
  p3 <- readPNG('p3.png')
  p4 <- readPNG('p4.png')
  p5 <- readPNG('p5.png')
  p6 <- readPNG('p6.png')
  p7 <- readPNG('p7.png')
  p8 <- readPNG('p8.png')
  p9 <- readPNG('p9.png')
  
  
  grid.arrange(rasterGrob(p1), rasterGrob(p2),rasterGrob(p3),rasterGrob(p4),rasterGrob(p5),
               rasterGrob(p6),rasterGrob(p7),rasterGrob(p8),rasterGrob(p9),ncol=2)][1]
  

  
• How to use map/reduce and read_csv to import a subset of columns from multiple files

  I need to import multiple files into R. From each file I only need a subset of columns. I have working code that will import ALL 97 columns from an arbitrary number of files, but cannot figure out how to limit the import to only the 4 columns I need.

  I rely on the tidyverse for many of my functions. R version: 3.6.0

  I have created code that imports limited columns from a SINGLE file but cannot expand this to all files.

  #this code works correctly and imports all columns from multiple files
  setwd('R:\\Data\\Raw Data Files')
  files <- c('686960.csv','686962.csv','686963.csv','623056.csv','608910.csv','608911.csv','609248.csv','609246.csv', '722178.csv', '722177.csv', '756542.csv')
  
  data <- files %>% 
  map(read_csv) %>% 
    reduce(rbind)
  
  #this code works correctly and imports 4 columns from 1 file
  test <- test_file %>%  read_csv(
                  col_types = 
                    cols_only(
                      'Member ID' = col_character()
                      ,'Patient Age' = col_character()
                      ,'Plan Payment Amount' = col_character()
                      ,'Check Date' = col_character()
                    )
                  ,n_max = 10
  )
  
  
  #this code produces an error
  setwd('R:\\Data\\Raw Data Files')
  files <- c('686960.csv','686962.csv','686963.csv','623056.csv','608910.csv','608911.csv','609248.csv','609246.csv', '722178.csv', '722177.csv', '756542.csv')
  
  data_test <- files %>% 
    map(read_csv(
      col_types = 
        cols_only(
          'Member ID' = col_character()
          ,'Patient Age' = col_character()
          ,'Plan Payment Amount' = col_character()
          ,'Check Date' = col_character()
        )
      ,n_max = 10
    )) %>% 
    reduce(rbind)
  
  #Error in read_delimited(file, tokenizer, col_names = col_names, col_types = #col_types,  : 
  #  argument "file" is missing, with no default
  
  
• How to make sure the user of a shiny app is using the right package versions in R

  Due to recent experience with several bugs created by updating packages, I wonder what the best approach is for the following problem:

  I currently provide a stand alone version so to say of my shiny App (just the script files to run it locally) and run a long list of require() functions to load / install the needed packages. However, in the end I would like to use fixed package versions to avoid bugs created by changes in packages.

  Is there a way to ensure that the user, who may have older or newer versions of packages on their computer, is using the right version of all the packages my app needs?
• Calculate plate stresses given displacements in Matlab

  Matlab has a convenient toolbox to calculate the displacements of isotropic plates subjected to uniformly distributed load. There's an example online from Matlab here.

  So, after a model has been completely defined, the program calculates displacements as:

  res = solvepde(model);
  

  the displacement is

  u = res.NodalSolution
  

  But there isn't an output that has the stresses sx, sy, sxx, syy, svonmises.

  res = 
  
    StationaryResults with properties:
  
      NodalSolution: [1893×2 double]
         XGradients: [1893×2 double]
         YGradients: [1893×2 double]
         ZGradients: [0×2 double]
               Mesh: [1×1 FEMesh]
  

  Any help on how can the stresses be calculated for any thickness t given the above inputs?

  Thanks!
• Recommendation for Polygon equipartitioning code

  Here is a code which takes in a rectilinear polygon and finds the maximum and minimum number of overlapping rectangular decomposition. I am interested in a python or Matlab code that takes in a rectilinear polygon and outputs n equal partitions.

  Some clarifications are needed. They are as follows.

  n is given by the user. User may want to decompose a polygon to 3 or 5 or 10 equal parts.

  Now when I say equal I of course do not rigidly demand exactly equal because such an algorithm would be too slow and too hard to find. Heuristic solvers that find approximately equal partitions also work fine.

  Partitions themselves can be rectilinear polygons or otherwise.

  And the original polygon also need not be rectilinear. It can be general. Just to make the search simple, I suggested keeping the polygons rectilinear.

  Can you suggest me an available piece of software code in python or Matlab that does this?

  If not, please suggest a complete algorithm from scratch and how to implement it?
• Function with sparse matrix and null matrix( zero lines and zero columns)

  Function with a variable type structure which needs to return the number of sparse matrix field and null matrix ( zero lines and 0 columns) and to use issparse function.The structure would be read from another script document.

  I tried a bunch of stuff but it didn't seemed to work, i'm new to Matlab since i worked just with C++.Thank you.
• CharSequnce interface in Java 11 added method `compare`. Why not `compareTo` of Comparable interface?

  The CharSequence interface gained a new method in Java 11: compare.

  This method returns an int:

  the value 0 if the two CharSequence are equal; a negative integer if the first CharSequence is lexicographically less than the second; or a positive integer if the first CharSequence is lexicographically greater than the second.

  That sounds just like compareTo of Comparable. Yet the Java team obviously chose to not make CharSequence extend Comparable. Why not? The logic escapes me.

  ➥ What is it about CharSequence::compare that would not be an appropriate fit for Comparable::compareTo?
• What's the best way to compare two questionnaires

  What's the best way to compare two questionnaires(same questions but different answers) then echo similarity in percentage using php/codeigniter
• How to compare objects inside two List's with only one loop?

  I have two lists of objects with the same type (Transaction), and i need to compare them. My goal is, first of all, compare "id" and secondly, if they matched, check if matches "amount".

  How can i do this without using double loop?

  Lists looks like this:

  List 1 - [Transaction{id=123, amount=100.05}, Transaction{id=124, amount=150.75}, Transaction{id=125, amount=1010.0}, Transaction{id=126, amount=15.5}]

  List 2 - [Transaction{id=123, amount=94.7}, Transaction{id=124, amount=150.75}, Transaction{id=125, amount=1020.2}, Transaction{id=126, amount=15.5}, Transaction{id=127, amount=120.74]

  public class Transaction {
  
      private Integer id;
  
      private Double amount;
  
      public Transaction(Integer id, Double amount) {
          this.id = id;
          this.amount = amount;
      }
  
      public Integer getId() {
          return id;
      }
      public String getData() {
          return data;
      }
    }
  

  Tried smth like this:

          for (Transaction transactionSql : service.findAll()) {
          for (Transaction transactionTxt : scraper.read()) {
              if(transactionSql.getId().equals(transactionTxt.getId()) {
                  // do smth
              }
  
          }
      }
  
• MATLAB C MEX API : fread directly into struct?

  In pure C, I can open a binary file and read the file header directly into a struct pretty simply. In MATLAB's C-MEX API, it seems to be a much more tedious procedure. I have to 1: Initialize the struct from an array of field names. 2: Initialize a typed mxArray for each struct field individually. 3: fread into the mxArray for each struct field individually. 4: Assign an mxArray to each struct field individually (by number or name).

  Is there a simpler way to fread data directly into a MATLAB struct?

  #pragma pack(2)  
  typedef struct file_header  
  {  
      int64_t preamble;  
      int16_t version;  
      int16_t mode;  
      int8_t software_name[32];  
      int32_t content;  
      // ...  
      // other fields  
  } file_header_t;  
  
  void mexFunction(int nlhs, mxArray *plhs[], int nrhs, const mxArray *prhs[])  
  {  
      char *file_name = mxArrayToString(prhs[0]);  
      // read file header (C) /////////////////////////////////////////////////  
      file_header_t das_header;  
      FILE *fptr = fopen(file_name, "rb");  
      fread(&das_header, sizeof(file_header_t), 1, fptr);  
      fclose(fptr);  
  
      // read file header (MATLAB) ////////////////////////////////////////////  
      const int n_field = 37;  
      static const char *field_name[] = {  
          "preamble",  
          "version",  
          "mode",  
          "software_name",  
          "content"  
          // ...  
          // other field names  
      };  
      // init matlab struct  
      plhs[0] = mxCreateStructMatrix(1, 1, n_field, field_name);  
      // init mxArray containers for struct fields  
      mxArray *preamble = mxCreateNumericMatrix(1, 1, mxINT64_CLASS, mxREAL);  
      mxArray *version = mxCreateNumericMatrix(1, 1, mxINT16_CLASS, mxREAL);  
      mxArray *mode = mxCreateNumericMatrix(1, 1, mxINT16_CLASS, mxREAL);  
      mxArray *software_name = mxCreateNumericMatrix(1, 32, mxINT8_CLASS, mxREAL);  
      mxArray *content = mxCreateNumericMatrix(1, 1, mxINT32_CLASS, mxREAL);  
      // ...  
      // mxArray *all *subsequent *struct *fields;  
  
      fptr = fopen(file_name, "rb");  
      fread((int64_t*)mxGetData(preamble), 8, 1, fptr);  
      fread((int16_t*)mxGetData(version), 2, 1, fptr);  
      fread((int16_t*)mxGetData(mode), 2, 1, fptr);  
      fread((int8_t*)mxGetData(software_name), 1, 32, fptr);  
      fread((int32_t*)mxGetData(content), 4, 1, fptr);  
      // ...  
      // fread((type*)mxGetData(field_name), sizeof(type), 1, fptr);  
  
      fclose(fptr);  
  
      // assign field values  
      mxSetFieldByNumber(plhs[0], 0, 0, preamble);  
      mxSetFieldByNumber(plhs[0], 0, 1, version);  
      mxSetFieldByNumber(plhs[0], 0, 2, mode);  
      mxSetFieldByNumber(plhs[0], 0, 3, software_name);  
      mxSetFieldByNumber(plhs[0], 0, 4, content);  
      // ...  
      // mxSetFieldByNumber(plhs[0], 0, k, field_name);  
  }  
  
  
• How to get function offset?

  How can I get function offset? I found MexScript code, which returns function offset from compiled .dll library.

  findloc OFFSET binary "\x10\x31\x00\x00\x00"
  math OFFSET + 5
  math OFFSET + 0xa0
  goto OFFSET
  get funcoff long
  math funcoff - 0x10000000
  print "Offset %funcoff|x%"
  

  Offset is 0x0001a6a0 which is good. So what is \x10\x31\x00\x00\x00? How does it work? How can I do this in C++?
• Invalid MEX-file: undefined symbol: fftwf_cleanup

  Invalid MEX-file
  '/home/tilakraj/Desktop/code/BPT/BPTTilakraj/NL-SAR-Toolbox/nlsartoolbox_v06/interfaces/mex/sarnlsar.mexa64':
  /home/tilakraj/Desktop/code/BPT/BPTTilakraj/NL-SAR-Toolbox/nlsartoolbox_v06/interfaces/mex/sarnlsar.mexa64:
  undefined symbol: fftwf_cleanup.