These are quite popular distributions among developers. It would be nice to have automated installation for them either.

Not that important but the last line of "ground_truth_package.h" should be:

#endif /* GROUND_TRUTH_PACKAGE_H_ */

instead of

#endif /* MEASUREMENT_PACKAGE_H_ */

Based on the "Docs/Input_Output File Format.txt" document, the input file line should have 8 members for Lidar and 9 members for Radar.

However, in "obj_pose-laser-radar-synthetic-input.txt", there are two extra members in each line.

What am I missing here?

The merged master with 'project_v2' failed to compile as reported in udacity forum. I tested in Windows 10 for VS2017 and Ubuntu in VirtualBox. And somebody also report the same problem in MacOS 10.12.4.
I believe somehow the current version of uWebSocket in github is not compatible.
https://discussions.udacity.com/t/udacity-ekf-repository-fails-to-build/248521

Hi @deniskrut, @andrewpaster, @baumanab
Notice that our project doesn't have license, that's a little unprofessional for a open source project, especially there has much extends adoption and development in an online education project. "No license terms" means users don't know it can be used freely. Another reference is Can I call my program "Open Source" even if I don't use an approved license?.

So here I suggest the project add license, recommend we choose one of the below license, firstly because they're all Permissive Licenses, but they have different redistribution disclaimer:

• MIT
• Apache-2.0
• BSD-3-Clause

Regards,
Brandon

Hi all, my computer is running on MacOS Sierra 10.12.5. I had two questions regarding to the compilation of the code and using cmake and make.

1. I just finish the code in /src and wanted to compile the code. I followed the instructions, creating a "build" folder, and ran cmake .., it gave me this:

-- Configuring done
-- Generating done
-- Build files have been written to: /Users/Work/Documents/udacity/CarND-Extended-Kalman-Filter-Project

Then I ran make, it gave me this:

make: *** No targets specified and no makefile found. Stop.

It seems like cmake was not able to find the Makefile in the upper directory of folder "build". What would possibly be the cause of the problem?

2. Because of the above error message I got, I tried to run make directly under the project root directory, it throughs me an error too:

[ 20%] Building CXX object CMakeFiles/ExtendedKF.dir/src/main.cpp.o
In file included from /Users/Work/Documents/udacity/CarND-Extended-Kalman-Filter-Project/src/main.cpp:3:0:
/Users/Work/Documents/udacity/CarND-Extended-Kalman-Filter-Project/src/json.hpp:67:10: error: #error "unsupported GCC version - see https://github.com/nlohmann/json#supported-compilers"
         #error "unsupported GCC version - see https://github.com/nlohmann/json#supported-compilers"
          ^
make[2]: *** [CMakeFiles/ExtendedKF.dir/src/main.cpp.o] Error 1
make[1]: *** [CMakeFiles/ExtendedKF.dir/all] Error 2
make: *** [all] Error 2

It seems like I am not having the right gcc version. I checked my gcc version by typing in gcc --version, and I got

gcc (GCC) 4.8.5
Copyright (C) 2015 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

I also installed gcc5, gcc6 and gcc7 but I don't know how to specify which version to use. Do I need to do so? I assume the make file contain the instructions to compile the C++ code for this particular project.

Sorry I am new to C++ and compiler. Thank you for your time.

CarND-Extended-Kalman-Filter-Project/data/obj_pose-laser-radar-synthetic-input.txt is not consistent. The velocity describes an oval, the velocity magnitude in x- and y-direction differs by an order of magnitude. The position describes a circle segment and the way back to the origin.

This hyperlink was shown in readme even if I am part of Nano degree program.

https://classroom.udacity.com/nanodegrees/nd013/parts/40f38239-66b6-46ec-ae68-03afd8a601c8/modules/0949fca6-b379-42af-a919-ee50aa304e6a/lessons/f758c44c-5e40-4e01-93b5-1a82aa4e044f/concepts/382ebfd6-1d55-4487-84a5-b6a5a4ba1e47

Getting an error when running make to get uWebSockets to install stating that the Localization file is not found. Unable to proceed.
Assistance appreciated.

Without any coverage in the course material leading to this project, we probably shouldn't assume every student knows what below means.

cmake >= v3.5
make >= v4.1
gcc/g++ >= v5.4

Dear, I have a query, I know that this is due to an error in the dimensions of the matrices, but when compiling it does not present any error, I attach the complete code and the instructions that generate the error.

          //IMPLEMENTACION DE UN ALGORITMO DE  APRENDIZAJE SOM-RNA KOHONEN EN C++

//El proceso de aprendizaje de SOM es estocástico, fuera de línea y no supervisado.
//clase VectorXd representa un vector de tamaño dinámico en Eigen.
//En este caso, el vector se inicializa con valores aleatorios usando la función Random.

//El código dentro de la macro eigen_assert(aLhs.rows() == aRhs.rows() && aLhs.cols() == aRhs.cols())
//comprueba si el número de filas y columnas de ambas matrices es el mismo.
//Si estas dimensiones no son iguales, se generará un error en tiempo de compilación.
#include
#include <Eigen/Dense>//Eigen es una biblioteca de C++ para álgebra lineal que proporciona tipos de datos y operaciones eficientes para trabajar con matrices y vectores.
#include
#include
using namespace Eigen;
using namespace std;

class KohonenNetwork{
public:
MatrixXd weights;//MatrixXd: Este es un tipo de datos proporcionado por la biblioteca Eigen
//que representa una matriz de valores de tipo double.
int inputSize;
int outputSize;

KohonenNetwork(int inputSize, int outputSize)
    : inputSize(inputSize), outputSize(outputSize) {
    // Inicialización aleatoria de pesos
    std::srand(std::time(0));//srand(): Es una función en C que inicializa la semilla del generador de números aleatorios utilizado por la función rand().

    weights = MatrixXd::Random(outputSize, inputSize);//VARIABLE PESOS

    // ::Random(outputSize, inputSize) : Este es un método que se llama en el tipo de datos MatrixXd
    //para inicializar la matriz con valores aleatorios.outputSize y inputSize son los tamaños de la matriz resultante, donde outputSize es el número de filas y inputSize es el número de columnas.
}

int findWinner(const VectorXd& input) const {//FUNCION PARA DETERMINAR NEURONA GANADORA
    int winner = 0;
    double minDistance = (weights.row(0) - input).squaredNorm();

    for (int i = 1; i < outputSize; ++i) {
        double distance = (weights.row(i) - input).squaredNorm();
        if (distance < minDistance) {
            minDistance = distance;
            winner = i;//neurona ganadora
        }
    }

    return winner;    }

/* int encontrarGanadora(const VectorXd& entrada) const {
// Encontrar la neurona ganadora (neurona con el peso más cercano a la entrada)
double mejorDistancia = std::numeric_limits::max();
int neuronaGanadora = 0;

    for (int i = 0; i < numNeuronas; ++i) {
        double distancia = (pesos.row(i) - entrada).norm();

        if (distancia < mejorDistancia) {
            mejorDistancia = distancia;
            neuronaGanadora = i;//NEURONA GANADORA
        }
    }

    return neuronaGanadora;
}

*/

void updateWeights(int winner, const VectorXd& input, double learningRate) {
    weights.row(winner) += learningRate * (input - weights.row(winner));
}

MatrixXd getWeights() const {
    return weights;
}

};

int main() {
// Configuración de la red
const int inputSize = 2;
const int outputSize = 4;
const int numTrainingSamples = 1000;
const double learningRate = 0.1;
int winner;
const int numSensores = 4;
// VectorXd input;

// Inicialización de la semilla para números aleatorios
srand(static_cast<unsigned int>(time(nullptr)));



/*entrada << -1, 0,
           -1, 1,
           -1, -1,
            0, -1,
            0, 1,
            0, 0,
            1, 1,
            1, -1,
            1, 0;
std::cout <<entrada;*/

            
/* ([[-1, 0], // sin obstaculos
    [-1, 1], //sin obstaculos
    [-1, -1], // sin obstaculos
    [0, -1], // obstaculo detectado a derecha
    [0, 1], // obstaculo a izq
    [0, 0], //obstaculo centro
    [1, 1],// demasiado cerca a derecha
    [1, -1], //demasiado cerca a izq
    [1, 0]    // demasiado cerca centro
] ); */
// Crear datos de entrenamiento (simulados)
MatrixXf trainingData(inputSize, numTrainingSamples);
trainingData = MatrixXf::Random(inputSize, numTrainingSamples);

        // Crear y entrenar la red
     KohonenNetwork kohonen(inputSize, outputSize);

     // Entrada de ejemplo desde los sensores (valores digitales aleatorios)
   
     VectorXf entrada = VectorXf::Random(numSensores);//En este código, VectorXd es un tipo de datos proporcionado por la biblioteca Eigen
   
     
     //Eigen::MatrixXd m(2, 2);  // Entrada de ejemplo desde los sensores (valores digitales aleatorios)
  //   m(0, 0) = 3;
  //   m(1, 0) = 2.5;
  //   m(0, 1) = -1;
   //  m(1, 1) = 2;
     //para representar vectores dinámicos de tamaño variable. 
     //La función Random genera valores aleatorios en el rango [-1, 1] y los asigna al vector entrada. Este vector ahora contiene numEntradas valores digitales aleatorios.
    
for (int epoch = 0; epoch < 1000; ++epoch) {
    // Realizar un pase hacia adelante y hacia atrás por cada ejemplo de entrenamiento
    for (int i = 0; i < numTrainingSamples; ++i) {
        VectorXf input = trainingData.col(i);

        // Pase hacia adelante y hacia atrás
     winner = kohonen.findWinner(input);//genera error
      int neuronaGanadora =kohonen.findWinner(entrada);//GENERARA ERROR
     kohonen.updateWeights(winner, input, learningRate);//genera error
    }
}
// Obtener los pesos finales
MatrixXd finalWeights = kohonen.getWeights();
// Mostrar resultados

// std::cout << "datos de entrenamiento: \n" << trainingData.transpose() << "\n";
std::cout << "Pesos finales de la red de Kohonen:\n" << finalWeights << "\n";
std::cout << "datos de entrada de la red de Kohonen:\n" << entrada << "\n";
std::cout << "datos de entrenamiento:\n " <<trainingData << "\n";
std::cout << "neurona de salida:\n" << outputSize << "\n";
std::cout << "neurona de salida:\n" << inputSize << "\n";
// std::cout << "mostramos el vector de entrada m:\n" << m << std::endl;

}

The institutions that generate this error are the following:

// Pase hacia adelante y hacia atrás
winner = kohonen.findWinner(input);//genera error
int neuronaGanadora =kohonen.findWinner(entrada);//GENERARA ERROR
kohonen.updateWeights(winner, input, learningRate);//genera error

Thank you for any suggestions to solve the problem during the compilation of the code developed in C++

I had been using Eigen with another application and noticed a recurring seg fault when trying to end the application. valgrind pointed me to a memory alignment issue within Eigen, which led me to this:

https://eigen.tuxfamily.org/dox/group__TopicStructHavingEigenMembers.html

I believe that your code for all term 2 projects that use Eigen needs to include EIGEN_MAKE_ALIGNED_OPERATOR_NEW in the public members area wherever an Eigen container is declared within a struct/class.

Hello,

I am using this code for experiments for other synthetic data where data is generated by matlab , I see that estimations are totally wrong when the (x, y) both are negative which means the when phi is 180 to 270. Can you please let me know if you have any idea ?

Regards
Bheeshma

Basic Build Instructions

Clone this repo.
Make a build directory: mkdir build && cd build
Compile: cmake .. && make
On windows, you may need to run: cmake .. -G "Unix Makefiles" && make
Run it: ./ExtendedKF path/to/input.txt path/to/output.txt. You can find some sample inputs in 'data/'.
eg. ./ExtendedKF ../data/obj_pose-laser-radar-synthetic-input.txt

The install-mac.sh fails and produces the following error:

CMake Error at /usr/local/Cellar/cmake/3.8.2/share/cmake/Modules/FindPackageHandleStandardArgs.cmake:137 (message): Could NOT find OpenSSL, try to set the path to OpenSSL root folder in the system variable OPENSSL_ROOT_DIR (missing: OPENSSL_INCLUDE_DIR) Call Stack (most recent call first): /usr/local/Cellar/cmake/3.8.2/share/cmake/Modules/FindPackageHandleStandardArgs.cmake:377 (_FPHSA_FAILURE_MESSAGE) /usr/local/Cellar/cmake/3.8.2/share/cmake/Modules/FindOpenSSL.cmake:387 (find_package_handle_standard_args) CMakeLists.txt:8 (find_package)

I had to change cmake .. with cmake -DOPENSSL_ROOT_DIR=/usr/local/Cellar/openssl/1.0.2l -DOPENSSL_LIBRARIES=/usr/local/Cellar/openssl/1.0.2l/lib .. to make the script work as mentioned in #32 .

Can you adjust the script that it automatically detects the openssl root dir?

I've just tried to install uWebSocket on Mac OS Sierra using install-mac.sh and got the following error:

CMake Error at cmake_install.cmake:39 (file):
   file cannot create directory: /usr/lib64.  Maybe need administrative
   privileges.

make: *** [install] Error 1

Apparently the System Integrity Protection introduced in Mac OS 10.11 prohibits even super users from making changes to /usr.

I tried to change the cmake .. line to

cmake .. -DOPENSSL_ROOT_DIR=/usr/local/opt/openssl -DOPENSSL_LIBRARIES=/usr/local/opt/openssl/lib,

but I am still getting the same error.

gcc>=4.5 not 5.4 as given as a pre-requisite in README.md

Whenever restarting or switching datasets (which includes a restart) in the simulator environment,
the kalman filter object retains its object state (state matrix, process covariance matrix, etc), although the object should be re-initialized.

• Not reinitializing them actually makes the performance worse on a second run as compared to a first run
• Not reinitializing upon switching environments can actually make the kalman filter code crash

I printed out the full messages being received from the websocket connection, but there doesnt seem to be one including a flag for either 'restart' or for 'switched datasets'.

Currently seemingly only solution is upon restart, also quit the Kalman filter code and restart the code, but this is cumbersome. Is there a way to make the kalman filter code more robust by receiving some event when the code has to be re-initialized?

When I cloned the repo and did a test run, I got a "Segmentation fault: 11" seemingly near the end of the program. Is this is expected?

Terminal output:
Jamess-iMac:build Jimmy$ cmake .. && make
-- Configuring done
-- Generating done
-- Build files have been written to: /Users/Jimmy/Documents/Udacity/CarND-Extended-Kalman-Filter-Project/build
Scanning dependencies of target ExtendedKF
[ 20%] Building CXX object CMakeFiles/ExtendedKF.dir/src/tools.cpp.o
[ 40%] Linking CXX executable ExtendedKF
[100%] Built target ExtendedKF
Jamess-iMac:build Jimmy$ ./ExtendedKF ../data/sample-laser-radar-measurement-data-1.txt output.txt
EKF:
x_ = 1
1
1
1
P_ =
...
x_ = 1
1
1
1
P_ =
Accuracy - RMSE:
Segmentation fault: 11
Jamess-iMac:build Jimmy$

Hi after running cmake .. I have issue running make command：

D:\Git_repo\CarND-Extended-Kalman-Filter-Project\build>cmake ..
-- Building for: Visual Studio 14 2015
-- The C compiler identification is MSVC 19.0.24215.1
-- The CXX compiler identification is MSVC 19.0.24215.1
-- Check for working C compiler: C:/Program Files (x86)/Microsoft Visual Studio 14.0/VC/bin/cl.exe
-- Check for working C compiler: C:/Program Files (x86)/Microsoft Visual Studio 14.0/VC/bin/cl.exe -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working CXX compiler: C:/Program Files (x86)/Microsoft Visual Studio 14.0/VC/bin/cl.exe
-- Check for working CXX compiler: C:/Program Files (x86)/Microsoft Visual Studio 14.0/VC/bin/cl.exe -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Configuring done
-- Generating done
-- Build files have been written to: D:/Git_repo/CarND-Extended-Kalman-Filter-Project/build

D:\Git_repo\CarND-Extended-Kalman-Filter-Project\build>make
make: *** No targets specified and no makefile found.  Stop.

hello, i make successed and run ./ExtendedKF in build folder,
and the output is: Listening to port 4567, don't see any output.

libuv1-dev can't be installed using apt on ubuntu 14.04, which causes 'install_ubuntu.sh' to fail.
The bash script can be fixed by downloading the .deb(https://packages.ubuntu.com/xenial/amd64/libuv1-dev/download) and installing it (in case of 14.04).

I noticed timestamp_ in GroundTruthPackage class uses long (32bits) but in MeasurementPackage it uses long long (64 bits). Shouldn't we use same number of bits in both places? And it seems 32 bits cannot hold all widths of time stamps.

It probably won't report out error as we care only about delta in time stamps, but it will incur truncation by long(32bits) I believe.

Here is the error that i got:

cala-am02024907:build lauc136$ make
Scanning dependencies of target ExtendedKF
[ 20%] Building CXX object CMakeFiles/ExtendedKF.dir/src/main.cpp.o
/Users/lauc136/Downloads/CarND-Extended-Kalman-Filter-Project-project_v2/src/main.cpp:41:5: error:
no matching member function for call to 'onMessage'
h.onMessage([&fusionEKF,&tools,&estimations,&ground_truth](uWS::WebSoc...
~~^~~~~~~~~
/usr/local/include/uWS/Group.h:69:10: note: candidate function not viable: no
known conversion from '(lambda at
/Users/lauc136/Downloads/CarND-Extended-Kalman-Filter-Project-project_v2/src/main.cpp:41:15)'
to 'std::function<void (WebSocket *, char *, size_t, OpCode)>' for
1st argument
void onMessage(std::function<void(WebSocket *, char *, siz...
^
/usr/local/include/uWS/Group.h:69:10: note: candidate function not viable: no
known conversion from '(lambda at
/Users/lauc136/Downloads/CarND-Extended-Kalman-Filter-Project-project_v2/src/main.cpp:41:15)'
to 'std::function<void (WebSocket *, char *, size_t, OpCode)>' for
1st argument
void onMessage(std::function<void(WebSocket *, char *, siz...
^
/Users/lauc136/Downloads/CarND-Extended-Kalman-Filter-Project-project_v2/src/main.cpp:165:5: error:
no matching member function for call to 'onConnection'
h.onConnection([&h](uWS::WebSocketuWS::SERVER ws, uWS::HttpRequest req) {
~~^~~~~~~~~~~~
/usr/local/include/uWS/Group.h:67:10: note: candidate function not viable: no
known conversion from '(lambda at
/Users/lauc136/Downloads/CarND-Extended-Kalman-Filter-Project-project_v2/src/main.cpp:165:18)'
to 'std::function<void (WebSocket *, HttpRequest)>' for 1st argument
void onConnection(std::function<void(WebSocket *, HttpRequ...
^
/usr/local/include/uWS/Group.h:67:10: note: candidate function not viable: no
known conversion from '(lambda at
/Users/lauc136/Downloads/CarND-Extended-Kalman-Filter-Project-project_v2/src/main.cpp:165:18)'
to 'std::function<void (WebSocket *, HttpRequest)>' for 1st
argument
void onConnection(std::function<void(WebSocket *, HttpRequ...
^
/Users/lauc136/Downloads/CarND-Extended-Kalman-Filter-Project-project_v2/src/main.cpp:169:5: error:
no matching member function for call to 'onDisconnection'
h.onDisconnection([&h](uWS::WebSocketuWS::SERVER ws, int code, char ...
~~^~~~~~~~~~~~~~~
/usr/local/include/uWS/Group.h:70:10: note: candidate function not viable: no
known conversion from '(lambda at
/Users/lauc136/Downloads/CarND-Extended-Kalman-Filter-Project-project_v2/src/main.cpp:169:21)'
to 'std::function<void (WebSocket *, int, char *, size_t)>' for 1st
argument
void onDisconnection(std::function<void(WebSocket *, int c...
^
/usr/local/include/uWS/Group.h:70:10: note: candidate function not viable: no
known conversion from '(lambda at
/Users/lauc136/Downloads/CarND-Extended-Kalman-Filter-Project-project_v2/src/main.cpp:169:21)'
to 'std::function<void (WebSocket *, int, char *, size_t)>' for 1st
argument
void onDisconnection(std::function<void(WebSocket *, int c...
^
3 errors generated.
make[2]: *** [CMakeFiles/ExtendedKF.dir/src/main.cpp.o] Error 1
make[1]: *** [CMakeFiles/ExtendedKF.dir/all] Error 2
make: *** [all] Error 2
cala-am02024907:build lauc136$

I try to run this project in Qt Creator,then run term2_sim.x86_64,there is a error lead to unexpectedly finished of the program.
Error show:
Starting /home/xiefeng/Udacity/CarND-Extended-Kalman-Filter-Project-master/build/ExtendedKF...
Listening to port 4567
Connected!!!
EKF:
ExtendedKF: /home/xiefeng/Udacity/CarND-Extended-Kalman-Filter-Project-master/src/Eigen/src/Core/CwiseBinaryOp.h:132: Eigen::CwiseBinaryOp<BinaryOp, Lhs, Rhs>::CwiseBinaryOp(const Lhs&, const Rhs&, const BinaryOp&) [with BinaryOp = Eigen::internal::scalar_difference_op; Lhs = const Eigen::Matrix<double, -1, 1>; Rhs = const Eigen::GeneralProduct<Eigen::Matrix<double, -1, -1>, Eigen::Matrix<double, -1, 1>, 4>]: Assertion `aLhs.rows() == aRhs.rows() && aLhs.cols() == aRhs.cols()' failed.
The program has unexpectedly finished.

Hi the link you provided for make for windows is 3.8.1 but the minimum requirement is 4.1. I have run the model on both windows and ubuntu and see very different results. Not sure if it is the make issue here.
https://discussions.udacity.com/t/large-discrepancy-in-results-running-on-windows-and-ubuntu/229811

Running the project using

./ExtendedKF ../data/sample-laser-radar-measurement-data-1.txt output.txt > error.txt

Gives me the following error

[1] 27390 segmentation fault ./ExtendedKF ../data/sample-laser-radar-measurement-data-1.txt output.txt >

(This error appears without piping to a file as well, but I piped to a file so I could attach the output to this post)

OS X version 10.12.3
CMake 3.7.2
Make 3.81 (I couldn't find a way to update OS X to Make 4.1)

And attached are the relevant files
output.txt
error.txt
sample-laser-radar-measurement-data-1.txt

I was setting up the environment for the EKF project. I installed the Ubuntu bash on Windows 10, and I was trying to run the simulator.

First I extracted the simulator to a folder, changed the permission to execute the file
chmod ug+x ./term2_sim.x86_64

Then I found that in order to run the exe there are some dependencies, so I installed all the dependencies it complained for
sudo apt-get install libglu1
sudo apt-get install libxcursor1
sudo apt-get install libxrandr2

But now wen I run this command to run the simulator.
sudo ./term2_sim.x86_64

I get this error

"Segmentation fault (core dumped)"

I am not sure if I am doing something wrong, could you please help ?

./install-ubuntu.sh does not complete unless the Ubuntu version is 16.04. I came across complaints on the forum that versions older than 16.04 face similar issues but even a newer version (18) did not succeed.