Build Error

[J-J-Wang]

Hi,
I encountered the following compilation issue while using fgm_code:
/fgm-cpp-0.0.2/fgm.cpp:3: /usr/include/Eigen/src/Core/CwiseBinaryOp.h:48:8: error: incomplete type ‘Eigen::internal::cwise_promote_storage_order<Eigen::Sparse, Eigen::Sparse, 0, 1>’ used in nested name specifier
and cmake info as follow:
`Make Deprecation Warning at CMakeLists.txt:1 (cmake_minimum_required):
Compatibility with CMake < 2.8.12 will be removed from a future version of
CMake.

Update the VERSION argument value or use a ... suffix to tell
CMake that the project does not need compatibility with older versions.

-- The C compiler identification is GNU 9.4.0
-- The CXX compiler identification is GNU 9.4.0
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working C compiler: /usr/bin/cc - skipped
-- Detecting C compile features
-- Detecting C compile features - done
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Check for working CXX compiler: /usr/bin/c++ - skipped
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Found PythonInterp: /usr/bin/python (found version "2.7.17")
-- Found PythonLibs: /usr/lib/x86_64-linux-gnu/libpython2.7.so (found version "2.7.17")
-- Found Boost: /usr/include (found version "1.65.1") found components: python
-- Configuring done
-- Generating done
-- Build files have been written to: /home/cidi/Downloads/fgm-cpp-0.0.2/build
`
Another question
function fgm(Eigen::MatrixXd& KP, Eigen::MatrixXd& KQ,
Eigen::MatrixXd& Ct, Eigen::MatrixXd& asgTX,
std::map<std::string, Eigen::MatrixXd>& gph1,
std::map<std::string, Eigen::MatrixXd>& gph2,
int nAlp = 101, int nItMa = 100, int nHst = 10)；

What do Ct and asgTX matrix mean?

Looking forward to your reply.

BR

[WateenGQY]

build error has been sloved.

• use SparseMatrix<double, Eigen::RowMajor> G1s = G1.sparseView(); instead of SparseMatrix<double> G1s = G1.sparseView();
• use temporary variable like

SparseMatrix<double, Eigen::RowMajor> G1s_adjoint = G1s.adjoint();  
GXGs = G1s_adjoint * Xs * G2s;

instead of

GXGs = G1s.adjoint() * Xs * G2s;

There are many similarities that need to be modified

Eigen3_VERSION:3.3.7

[WateenGQY]

I guess that
asgTX is the groundtruth of X.

Ct seems like the initial value of X by {0,1} but dimensions are not right.

and I want to confirm what is gph1["H"]?

    MatrixXd G1 = gph1["G"];
    MatrixXd H1 = gph1["H"];
    MatrixXd G2 = gph2["G"];
    MatrixXd H2 = gph2["H"];

[J-J-Wang]

I guess that asgTX is the groundtruth of X.

Ct seems like the initial value of X by {0,1} but dimensions are not right.

and I want to confirm what is gph1["H"]?

    MatrixXd G1 = gph1["G"];
    MatrixXd H1 = gph1["H"];
    MatrixXd G2 = gph2["G"];
    MatrixXd H2 = gph2["H"];

Thank you reply, this mean you can read paper:Feng Zhou, and F. De la Torre. “Factorized Graph Matching.” 2012 IEEE Conference on Computer Vision and Pattern Recognition, 2012, https://doi.org/10.1109/cvpr.2012.6247667. asgTX maybe is the groundtruth of X, but 'Ct' is not appear in paper.

[WateenGQY]

I guess that asgTX is the groundtruth of X.
Ct seems like the initial value of X by {0,1} but dimensions are not right.
and I want to confirm what is gph1["H"]?

    MatrixXd G1 = gph1["G"];
    MatrixXd H1 = gph1["H"];
    MatrixXd G2 = gph2["G"];
    MatrixXd H2 = gph2["H"];

Thank you reply, this mean you can read paper:Feng Zhou, and F. De la Torre. “Factorized Graph Matching.” 2012 IEEE Conference on Computer Vision and Pattern Recognition, 2012, https://doi.org/10.1109/cvpr.2012.6247667. asgTX maybe is the groundtruth of X, but 'Ct' is not appear in paper.

thanks for your reply. according to the paper, $$H_1 = [G_1, I_{n1}]$$ has dimension n1*(m1+n1) but it is resized for n2*m1. I am a little confused.

    if (n1 < n2) {
        auto mi = KP.minCoeff();
        KP = resize(KP, n2, n2, mi);
        G1 = resize(G1, n2, m1, 0);
        H1 = resize(H1, n2, m1, 0);
        Ct = resize(Ct, n2, n2, 1);
    } else if (n1 > n2) {
        auto mi = KP.minCoeff();
        KP = resize(KP, n1, n1, mi);
        G2 = resize(G2, n1, m2, 0);
        H2 = resize(H2, n1, m2, 0);
        Ct = resize(Ct, n1, n1, 1);
    }

hi, 我正好也用到这个，我们方便留个联系方式交流一下么？

[J-J-Wang]

I guess that asgTX is the groundtruth of X.
Ct seems like the initial value of X by {0,1} but dimensions are not right.
and I want to confirm what is gph1["H"]?

    MatrixXd G1 = gph1["G"];
    MatrixXd H1 = gph1["H"];
    MatrixXd G2 = gph2["G"];
    MatrixXd H2 = gph2["H"];

Thank you reply, this mean you can read paper:Feng Zhou, and F. De la Torre. “Factorized Graph Matching.” 2012 IEEE Conference on Computer Vision and Pattern Recognition, 2012, https://doi.org/10.1109/cvpr.2012.6247667. asgTX maybe is the groundtruth of X, but 'Ct' is not appear in paper.

thanks for your reply. according to the paper, H1=[G1,In1] has dimension n1*(m1+n1) but it is resized for n2*m1. I am a little confused.

    if (n1 < n2) {
        auto mi = KP.minCoeff();
        KP = resize(KP, n2, n2, mi);
        G1 = resize(G1, n2, m1, 0);
        H1 = resize(H1, n2, m1, 0);
        Ct = resize(Ct, n2, n2, 1);
    } else if (n1 > n2) {
        auto mi = KP.minCoeff();
        KP = resize(KP, n1, n1, mi);
        G2 = resize(G2, n1, m2, 0);
        H2 = resize(H2, n1, m2, 0);
        Ct = resize(Ct, n1, n1, 1);
    }

hi, 我正好也用到这个，我们方便留个联系方式交流一下么？

I guess that asgTX is the groundtruth of X.
Ct seems like the initial value of X by {0,1} but dimensions are not right.
and I want to confirm what is gph1["H"]?

    MatrixXd G1 = gph1["G"];
    MatrixXd H1 = gph1["H"];
    MatrixXd G2 = gph2["G"];
    MatrixXd H2 = gph2["H"];

Thank you reply, this mean you can read paper:Feng Zhou, and F. De la Torre. “Factorized Graph Matching.” 2012 IEEE Conference on Computer Vision and Pattern Recognition, 2012, https://doi.org/10.1109/cvpr.2012.6247667. asgTX maybe is the groundtruth of X, but 'Ct' is not appear in paper.

thanks for your reply. according to the paper, H1=[G1,In1] has dimension n1*(m1+n1) but it is resized for n2*m1. I am a little confused.

    if (n1 < n2) {
        auto mi = KP.minCoeff();
        KP = resize(KP, n2, n2, mi);
        G1 = resize(G1, n2, m1, 0);
        H1 = resize(H1, n2, m1, 0);
        Ct = resize(Ct, n2, n2, 1);
    } else if (n1 > n2) {
        auto mi = KP.minCoeff();
        KP = resize(KP, n1, n1, mi);
        G2 = resize(G2, n1, m2, 0);
        H2 = resize(H2, n1, m2, 0);
        Ct = resize(Ct, n1, n1, 1);
    }

hi, 我正好也用到这个，我们方便留个联系方式交流一下么？

+v: WangJun

[J-J-Wang]

I guess that asgTX is the groundtruth of X.
Ct seems like the initial value of X by {0,1} but dimensions are not right.
and I want to confirm what is gph1["H"]?

    MatrixXd G1 = gph1["G"];
    MatrixXd H1 = gph1["H"];
    MatrixXd G2 = gph2["G"];
    MatrixXd H2 = gph2["H"];

Thank you reply, this mean you can read paper:Feng Zhou, and F. De la Torre. “Factorized Graph Matching.” 2012 IEEE Conference on Computer Vision and Pattern Recognition, 2012, https://doi.org/10.1109/cvpr.2012.6247667. asgTX maybe is the groundtruth of X, but 'Ct' is not appear in paper.

thanks for your reply. according to the paper, H1=[G1,In1] has dimension n1*(m1+n1) but it is resized for n2*m1. I am a little confused.

    if (n1 < n2) {
        auto mi = KP.minCoeff();
        KP = resize(KP, n2, n2, mi);
        G1 = resize(G1, n2, m1, 0);
        H1 = resize(H1, n2, m1, 0);
        Ct = resize(Ct, n2, n2, 1);
    } else if (n1 > n2) {
        auto mi = KP.minCoeff();
        KP = resize(KP, n1, n1, mi);
        G2 = resize(G2, n1, m2, 0);
        H2 = resize(H2, n1, m2, 0);
        Ct = resize(Ct, n1, n1, 1);
    }

hi, 我正好也用到这个，我们方便留个联系方式交流一下么？

I guess that asgTX is the groundtruth of X.
Ct seems like the initial value of X by {0,1} but dimensions are not right.
and I want to confirm what is gph1["H"]?

    MatrixXd G1 = gph1["G"];
    MatrixXd H1 = gph1["H"];
    MatrixXd G2 = gph2["G"];
    MatrixXd H2 = gph2["H"];

Thank you reply, this mean you can read paper:Feng Zhou, and F. De la Torre. “Factorized Graph Matching.” 2012 IEEE Conference on Computer Vision and Pattern Recognition, 2012, https://doi.org/10.1109/cvpr.2012.6247667. asgTX maybe is the groundtruth of X, but 'Ct' is not appear in paper.

thanks for your reply. according to the paper, H1=[G1,In1] has dimension n1*(m1+n1) but it is resized for n2*m1. I am a little confused.

    if (n1 < n2) {
        auto mi = KP.minCoeff();
        KP = resize(KP, n2, n2, mi);
        G1 = resize(G1, n2, m1, 0);
        H1 = resize(H1, n2, m1, 0);
        Ct = resize(Ct, n2, n2, 1);
    } else if (n1 > n2) {
        auto mi = KP.minCoeff();
        KP = resize(KP, n1, n1, mi);
        G2 = resize(G2, n1, m2, 0);
        H2 = resize(H2, n1, m2, 0);
        Ct = resize(Ct, n1, n1, 1);
    }

hi, 我正好也用到这个，我们方便留个联系方式交流一下么？

+v: WangJun
前后各一个 下划线 _ ，这个没显示