Asymmetric distribution, Gauss curve











up vote
5
down vote

favorite
1












enter image description hereI want to create a positive and negative asymmetric distribution, as shown in the image, it will be possible to include the data (values) one by one to give the desired curve.



The WME is



documentclass[border=5mm]{standalone}
usepackage{pgfplots}

begin{document}

newcommandgauss[2]{1/(#2*sqrt(2*pi))*exp(-((x-#1)^2)/(2*#2^2))}
begin{tikzpicture}[
every pin edge/.style={latex-,line width=1.5pt},
every pin/.style={fill=yellow!50,rectangle,rounded corners=3pt,font=small}]
begin{axis}[every axis plot post/.append style={
mark=none,domain=-3.:3.,samples=100},
clip=false,
axis y line=none,
axis x line*=bottom,
ymin=0,
xtick=empty,]
addplot[line width=1.5pt,blue] {gauss{0.}{1.}};
node[pin=270:{$X=M_e=M_o$}] at (axis cs:0,0) {};
draw[line width=1.5pt,dashed, red] (axis description cs:0.5,0) -- (axis description cs:0.5,0.92);
end{axis}
end{tikzpicture}

end{document}









share|improve this question


























    up vote
    5
    down vote

    favorite
    1












    enter image description hereI want to create a positive and negative asymmetric distribution, as shown in the image, it will be possible to include the data (values) one by one to give the desired curve.



    The WME is



    documentclass[border=5mm]{standalone}
    usepackage{pgfplots}

    begin{document}

    newcommandgauss[2]{1/(#2*sqrt(2*pi))*exp(-((x-#1)^2)/(2*#2^2))}
    begin{tikzpicture}[
    every pin edge/.style={latex-,line width=1.5pt},
    every pin/.style={fill=yellow!50,rectangle,rounded corners=3pt,font=small}]
    begin{axis}[every axis plot post/.append style={
    mark=none,domain=-3.:3.,samples=100},
    clip=false,
    axis y line=none,
    axis x line*=bottom,
    ymin=0,
    xtick=empty,]
    addplot[line width=1.5pt,blue] {gauss{0.}{1.}};
    node[pin=270:{$X=M_e=M_o$}] at (axis cs:0,0) {};
    draw[line width=1.5pt,dashed, red] (axis description cs:0.5,0) -- (axis description cs:0.5,0.92);
    end{axis}
    end{tikzpicture}

    end{document}









    share|improve this question
























      up vote
      5
      down vote

      favorite
      1









      up vote
      5
      down vote

      favorite
      1






      1





      enter image description hereI want to create a positive and negative asymmetric distribution, as shown in the image, it will be possible to include the data (values) one by one to give the desired curve.



      The WME is



      documentclass[border=5mm]{standalone}
      usepackage{pgfplots}

      begin{document}

      newcommandgauss[2]{1/(#2*sqrt(2*pi))*exp(-((x-#1)^2)/(2*#2^2))}
      begin{tikzpicture}[
      every pin edge/.style={latex-,line width=1.5pt},
      every pin/.style={fill=yellow!50,rectangle,rounded corners=3pt,font=small}]
      begin{axis}[every axis plot post/.append style={
      mark=none,domain=-3.:3.,samples=100},
      clip=false,
      axis y line=none,
      axis x line*=bottom,
      ymin=0,
      xtick=empty,]
      addplot[line width=1.5pt,blue] {gauss{0.}{1.}};
      node[pin=270:{$X=M_e=M_o$}] at (axis cs:0,0) {};
      draw[line width=1.5pt,dashed, red] (axis description cs:0.5,0) -- (axis description cs:0.5,0.92);
      end{axis}
      end{tikzpicture}

      end{document}









      share|improve this question













      enter image description hereI want to create a positive and negative asymmetric distribution, as shown in the image, it will be possible to include the data (values) one by one to give the desired curve.



      The WME is



      documentclass[border=5mm]{standalone}
      usepackage{pgfplots}

      begin{document}

      newcommandgauss[2]{1/(#2*sqrt(2*pi))*exp(-((x-#1)^2)/(2*#2^2))}
      begin{tikzpicture}[
      every pin edge/.style={latex-,line width=1.5pt},
      every pin/.style={fill=yellow!50,rectangle,rounded corners=3pt,font=small}]
      begin{axis}[every axis plot post/.append style={
      mark=none,domain=-3.:3.,samples=100},
      clip=false,
      axis y line=none,
      axis x line*=bottom,
      ymin=0,
      xtick=empty,]
      addplot[line width=1.5pt,blue] {gauss{0.}{1.}};
      node[pin=270:{$X=M_e=M_o$}] at (axis cs:0,0) {};
      draw[line width=1.5pt,dashed, red] (axis description cs:0.5,0) -- (axis description cs:0.5,0.92);
      end{axis}
      end{tikzpicture}

      end{document}






      tikz-pgf gauss






      share|improve this question













      share|improve this question











      share|improve this question




      share|improve this question










      asked Nov 26 at 2:34









      Samuel Diaz

      28618




      28618






















          3 Answers
          3






          active

          oldest

          votes

















          up vote
          5
          down vote



          accepted










          This is only a partial answer since it is not clear to me what an asymmetric Gauss curve precisely is. This is more to discuss how to set this up in principle. So I am only going to discuss how to plot a deformed Gauss curve.



          To this end, I'd like to convince you to use declare function rather than the definition you use. In the example below, I am going to use



          declare function={Gauss(x,y,z,u)=1/(z*sqrt(2*pi))*exp(-((x-y+u*(x-y)*sign(x-y))^2)/(2*z^2));


          Here Gauss reduces to an ordinary Gaussian for u=0, where x is just the variable, y defines the location of the maximum and z the width. If you turn on a nontrivial u, the Gaussian will get deformed.



          documentclass[border=5mm]{standalone}
          usepackage{pgfplots}
          pgfplotsset{height=4cm,width=8cm,compat=1.16}
          begin{document}

          begin{tikzpicture}[font=sffamily,
          declare function={Gauss(x,y,z,u)=1/(z*sqrt(2*pi))*exp(-((x-y+u*(x-y)*sign(x-y))^2)/(2*z^2));},
          every pin edge/.style={latex-,line width=1.5pt},
          every pin/.style={fill=yellow!50,rectangle,rounded corners=3pt,font=small}]
          begin{axis}[
          every axis plot post/.append style={
          mark=none,samples=101},
          clip=false,
          axis y line=none,
          axis x line*=bottom,
          ymin=0,
          xtick=empty,]
          addplot[line width=1.5pt,blue,domain=-1:3] {Gauss(x,0,0.6,-0.4)};
          draw[line width=1.5pt,dashed, black] (0,0) -- (0,{Gauss(0,0,0.6,-0.4)});
          %node[pin=270:{$X=M_e=M_o$}] at (axis cs:0,0) {};
          draw[line width=1.5pt,dashed, red] (0.6,0) -- (0.6,{Gauss(0.6,0,0.6,-0.4)});
          draw[line width=1.5pt,dashed, red] (-0.6,0) -- (-0.6,{Gauss(-0.6,0,0.6,-0.4)});
          path (-0.6,0) coordinate (ML) (0.6,0) coordinate (MR) (0,0) coordinate (MM);
          end{axis}
          draw[latex-] (ML) to[out=-90,in=45] ++ (-0.6,-0.6) node[below left,inner
          sep=1pt]{$langle Xrangle-Delta$};
          draw[latex-] (MR) to[out=-90,in=135] ++ (0.6,-0.6) node[below right,inner
          sep=1pt]{$langle Xrangle+Delta$};
          draw[latex-] (MM) --++ (0,-0.6) node[below,inner
          sep=1pt]{$langle Xrangle$};
          end{tikzpicture}
          end{document}


          enter image description here






          share|improve this answer























          • @ marmot They are two separate graphs, one that leans to the right and the other to the left, known as negative skew and positive skew. en.wikipedia.org/wiki/Skewness
            – Samuel Diaz
            Nov 26 at 3:25












          • @SamuelDiaz Thanks for the link! But as far as I can see it does not really give you a unique parametrization of these deformed Gaussians, does it?
            – marmot
            Nov 26 at 3:59










          • Correct, you have to keep in mind that you meet that average < average < mode or average > median > fashion.
            – Samuel Diaz
            Nov 26 at 4:09












          • @SamuelDiaz I added a possible way how you could use this.
            – marmot
            Nov 26 at 4:33


















          up vote
          4
          down vote













          Probably slightly overkill and certainly not efficient, but you could try an approximate skew normal (central tendencies are omitted below):



          documentclass[border=5, tikz]{standalone}
          usetikzlibrary{math}
          usepackage{pgfplots}
          pgfplotsset{compat=1.14}
          tikzmath{%
          function h1(x, lx) { return (9*lx + 3*((lx)^2) + ((lx)^3)/3 + 9); };
          function h2(x, lx) { return (3*lx - ((lx)^3)/3 + 4); };
          function h3(x, lx) { return (9*lx - 3*((lx)^2) + ((lx)^3)/3 + 7); };
          function skewnorm(x, l) {
          x = (l < 0) ? -x : x;
          l = abs(l);
          e = exp(-(x^2)/2);
          return (l == 0) ? 1 / sqrt(2 * pi) * e: (
          (x < -3/l) ? 0 : (
          (x < -1/l) ? e / (8 * sqrt(2 * pi)) * h1(x, x*l) : (
          (x < 1/l) ? e / (4 * sqrt(2 * pi)) * h2(x, x*l) : (
          (x < 3/l) ? e / (8 * sqrt(2 * pi)) * h3(x, x*l) : (
          sqrt(2/pi) * e)))));
          };
          }
          begin{document}
          begin{tikzpicture}[line join=round, line cap=round]
          begin{axis}[
          width=4in, height=2in,
          every axis plot post/.append style={
          mark=none, domain=-3.5:3.5, samples=200, very thick
          },
          clip=false,
          axis y line=none,
          axis x line*=bottom,
          ymin=0, ymax=0.75,
          xtick=empty,]
          addplot[red] {skewnorm(x, -4)};
          addplot[green] {skewnorm(x, -2)};
          addplot[gray] {skewnorm(x, 0)};
          addplot[blue] {skewnorm(x, 2)};
          addplot[orange] {skewnorm(x, 4)};
          legend{$lambda=-4$,$lambda=-2$,$lambda=0$,$lambda=2$,$lambda=4$}
          end{axis}
          end{tikzpicture}
          end{document}


          enter image description here






          share|improve this answer




























            up vote
            2
            down vote













            Another possible way (apart from @marmot's answer) is to plot the skewed distribution function is to exploit the chi-square distribution.



            For instance:



            documentclass{standalone}
            usepackage{pgfplots}
            %https://en.wikipedia.org/wiki/Chi-squared_distribution
            %https://tex.stackexchange.com/questions/120441/plot-the-probability-density-function-of-the-gamma-distribution?rq=1
            % the second link gives the numerical approximation of gamma function
            begin{document}
            begin{tikzpicture}[
            declare function={gamma(z)=
            (2.506628274631*sqrt(1/z) + 0.20888568*(1/z)^(1.5) + 0.00870357*(1/z)^(2.5) - (174.2106599*(1/z)^(3.5))/25920 - (715.6423511*(1/z)^(4.5))/1244160)*exp((-ln(1/z)-1)*z);},
            declare function={chipdf(x,k) = x^(k/2-1)*exp(-x/2) / (2^(k/2)*gamma(k));}
            ]

            begin{axis}[
            axis lines=left,
            enlargelimits=upper,
            ]
            addplot [smooth, domain=0:20, blue] {chipdf(x,2)};
            addplot [smooth, domain=0:20, green] {chipdf(x,3)};
            addplot [smooth, domain=0:20, black] {chipdf(x,4)};
            addplot [smooth, domain=0:20, cyan] {chipdf(x,5)};
            addplot [smooth, domain=0:20, magenta] {chipdf(x,6)};
            end{axis}
            end{tikzpicture}
            end{document}


            which will give you:



            enter image description here



            Using this, you can include your mean-median-mode as lines in the plot.






            share|improve this answer





















              Your Answer








              StackExchange.ready(function() {
              var channelOptions = {
              tags: "".split(" "),
              id: "85"
              };
              initTagRenderer("".split(" "), "".split(" "), channelOptions);

              StackExchange.using("externalEditor", function() {
              // Have to fire editor after snippets, if snippets enabled
              if (StackExchange.settings.snippets.snippetsEnabled) {
              StackExchange.using("snippets", function() {
              createEditor();
              });
              }
              else {
              createEditor();
              }
              });

              function createEditor() {
              StackExchange.prepareEditor({
              heartbeatType: 'answer',
              convertImagesToLinks: false,
              noModals: true,
              showLowRepImageUploadWarning: true,
              reputationToPostImages: null,
              bindNavPrevention: true,
              postfix: "",
              imageUploader: {
              brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
              contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
              allowUrls: true
              },
              onDemand: true,
              discardSelector: ".discard-answer"
              ,immediatelyShowMarkdownHelp:true
              });


              }
              });














              draft saved

              draft discarded


















              StackExchange.ready(
              function () {
              StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2ftex.stackexchange.com%2fquestions%2f461758%2fasymmetric-distribution-gauss-curve%23new-answer', 'question_page');
              }
              );

              Post as a guest















              Required, but never shown

























              3 Answers
              3






              active

              oldest

              votes








              3 Answers
              3






              active

              oldest

              votes









              active

              oldest

              votes






              active

              oldest

              votes








              up vote
              5
              down vote



              accepted










              This is only a partial answer since it is not clear to me what an asymmetric Gauss curve precisely is. This is more to discuss how to set this up in principle. So I am only going to discuss how to plot a deformed Gauss curve.



              To this end, I'd like to convince you to use declare function rather than the definition you use. In the example below, I am going to use



              declare function={Gauss(x,y,z,u)=1/(z*sqrt(2*pi))*exp(-((x-y+u*(x-y)*sign(x-y))^2)/(2*z^2));


              Here Gauss reduces to an ordinary Gaussian for u=0, where x is just the variable, y defines the location of the maximum and z the width. If you turn on a nontrivial u, the Gaussian will get deformed.



              documentclass[border=5mm]{standalone}
              usepackage{pgfplots}
              pgfplotsset{height=4cm,width=8cm,compat=1.16}
              begin{document}

              begin{tikzpicture}[font=sffamily,
              declare function={Gauss(x,y,z,u)=1/(z*sqrt(2*pi))*exp(-((x-y+u*(x-y)*sign(x-y))^2)/(2*z^2));},
              every pin edge/.style={latex-,line width=1.5pt},
              every pin/.style={fill=yellow!50,rectangle,rounded corners=3pt,font=small}]
              begin{axis}[
              every axis plot post/.append style={
              mark=none,samples=101},
              clip=false,
              axis y line=none,
              axis x line*=bottom,
              ymin=0,
              xtick=empty,]
              addplot[line width=1.5pt,blue,domain=-1:3] {Gauss(x,0,0.6,-0.4)};
              draw[line width=1.5pt,dashed, black] (0,0) -- (0,{Gauss(0,0,0.6,-0.4)});
              %node[pin=270:{$X=M_e=M_o$}] at (axis cs:0,0) {};
              draw[line width=1.5pt,dashed, red] (0.6,0) -- (0.6,{Gauss(0.6,0,0.6,-0.4)});
              draw[line width=1.5pt,dashed, red] (-0.6,0) -- (-0.6,{Gauss(-0.6,0,0.6,-0.4)});
              path (-0.6,0) coordinate (ML) (0.6,0) coordinate (MR) (0,0) coordinate (MM);
              end{axis}
              draw[latex-] (ML) to[out=-90,in=45] ++ (-0.6,-0.6) node[below left,inner
              sep=1pt]{$langle Xrangle-Delta$};
              draw[latex-] (MR) to[out=-90,in=135] ++ (0.6,-0.6) node[below right,inner
              sep=1pt]{$langle Xrangle+Delta$};
              draw[latex-] (MM) --++ (0,-0.6) node[below,inner
              sep=1pt]{$langle Xrangle$};
              end{tikzpicture}
              end{document}


              enter image description here






              share|improve this answer























              • @ marmot They are two separate graphs, one that leans to the right and the other to the left, known as negative skew and positive skew. en.wikipedia.org/wiki/Skewness
                – Samuel Diaz
                Nov 26 at 3:25












              • @SamuelDiaz Thanks for the link! But as far as I can see it does not really give you a unique parametrization of these deformed Gaussians, does it?
                – marmot
                Nov 26 at 3:59










              • Correct, you have to keep in mind that you meet that average < average < mode or average > median > fashion.
                – Samuel Diaz
                Nov 26 at 4:09












              • @SamuelDiaz I added a possible way how you could use this.
                – marmot
                Nov 26 at 4:33















              up vote
              5
              down vote



              accepted










              This is only a partial answer since it is not clear to me what an asymmetric Gauss curve precisely is. This is more to discuss how to set this up in principle. So I am only going to discuss how to plot a deformed Gauss curve.



              To this end, I'd like to convince you to use declare function rather than the definition you use. In the example below, I am going to use



              declare function={Gauss(x,y,z,u)=1/(z*sqrt(2*pi))*exp(-((x-y+u*(x-y)*sign(x-y))^2)/(2*z^2));


              Here Gauss reduces to an ordinary Gaussian for u=0, where x is just the variable, y defines the location of the maximum and z the width. If you turn on a nontrivial u, the Gaussian will get deformed.



              documentclass[border=5mm]{standalone}
              usepackage{pgfplots}
              pgfplotsset{height=4cm,width=8cm,compat=1.16}
              begin{document}

              begin{tikzpicture}[font=sffamily,
              declare function={Gauss(x,y,z,u)=1/(z*sqrt(2*pi))*exp(-((x-y+u*(x-y)*sign(x-y))^2)/(2*z^2));},
              every pin edge/.style={latex-,line width=1.5pt},
              every pin/.style={fill=yellow!50,rectangle,rounded corners=3pt,font=small}]
              begin{axis}[
              every axis plot post/.append style={
              mark=none,samples=101},
              clip=false,
              axis y line=none,
              axis x line*=bottom,
              ymin=0,
              xtick=empty,]
              addplot[line width=1.5pt,blue,domain=-1:3] {Gauss(x,0,0.6,-0.4)};
              draw[line width=1.5pt,dashed, black] (0,0) -- (0,{Gauss(0,0,0.6,-0.4)});
              %node[pin=270:{$X=M_e=M_o$}] at (axis cs:0,0) {};
              draw[line width=1.5pt,dashed, red] (0.6,0) -- (0.6,{Gauss(0.6,0,0.6,-0.4)});
              draw[line width=1.5pt,dashed, red] (-0.6,0) -- (-0.6,{Gauss(-0.6,0,0.6,-0.4)});
              path (-0.6,0) coordinate (ML) (0.6,0) coordinate (MR) (0,0) coordinate (MM);
              end{axis}
              draw[latex-] (ML) to[out=-90,in=45] ++ (-0.6,-0.6) node[below left,inner
              sep=1pt]{$langle Xrangle-Delta$};
              draw[latex-] (MR) to[out=-90,in=135] ++ (0.6,-0.6) node[below right,inner
              sep=1pt]{$langle Xrangle+Delta$};
              draw[latex-] (MM) --++ (0,-0.6) node[below,inner
              sep=1pt]{$langle Xrangle$};
              end{tikzpicture}
              end{document}


              enter image description here






              share|improve this answer























              • @ marmot They are two separate graphs, one that leans to the right and the other to the left, known as negative skew and positive skew. en.wikipedia.org/wiki/Skewness
                – Samuel Diaz
                Nov 26 at 3:25












              • @SamuelDiaz Thanks for the link! But as far as I can see it does not really give you a unique parametrization of these deformed Gaussians, does it?
                – marmot
                Nov 26 at 3:59










              • Correct, you have to keep in mind that you meet that average < average < mode or average > median > fashion.
                – Samuel Diaz
                Nov 26 at 4:09












              • @SamuelDiaz I added a possible way how you could use this.
                – marmot
                Nov 26 at 4:33













              up vote
              5
              down vote



              accepted







              up vote
              5
              down vote



              accepted






              This is only a partial answer since it is not clear to me what an asymmetric Gauss curve precisely is. This is more to discuss how to set this up in principle. So I am only going to discuss how to plot a deformed Gauss curve.



              To this end, I'd like to convince you to use declare function rather than the definition you use. In the example below, I am going to use



              declare function={Gauss(x,y,z,u)=1/(z*sqrt(2*pi))*exp(-((x-y+u*(x-y)*sign(x-y))^2)/(2*z^2));


              Here Gauss reduces to an ordinary Gaussian for u=0, where x is just the variable, y defines the location of the maximum and z the width. If you turn on a nontrivial u, the Gaussian will get deformed.



              documentclass[border=5mm]{standalone}
              usepackage{pgfplots}
              pgfplotsset{height=4cm,width=8cm,compat=1.16}
              begin{document}

              begin{tikzpicture}[font=sffamily,
              declare function={Gauss(x,y,z,u)=1/(z*sqrt(2*pi))*exp(-((x-y+u*(x-y)*sign(x-y))^2)/(2*z^2));},
              every pin edge/.style={latex-,line width=1.5pt},
              every pin/.style={fill=yellow!50,rectangle,rounded corners=3pt,font=small}]
              begin{axis}[
              every axis plot post/.append style={
              mark=none,samples=101},
              clip=false,
              axis y line=none,
              axis x line*=bottom,
              ymin=0,
              xtick=empty,]
              addplot[line width=1.5pt,blue,domain=-1:3] {Gauss(x,0,0.6,-0.4)};
              draw[line width=1.5pt,dashed, black] (0,0) -- (0,{Gauss(0,0,0.6,-0.4)});
              %node[pin=270:{$X=M_e=M_o$}] at (axis cs:0,0) {};
              draw[line width=1.5pt,dashed, red] (0.6,0) -- (0.6,{Gauss(0.6,0,0.6,-0.4)});
              draw[line width=1.5pt,dashed, red] (-0.6,0) -- (-0.6,{Gauss(-0.6,0,0.6,-0.4)});
              path (-0.6,0) coordinate (ML) (0.6,0) coordinate (MR) (0,0) coordinate (MM);
              end{axis}
              draw[latex-] (ML) to[out=-90,in=45] ++ (-0.6,-0.6) node[below left,inner
              sep=1pt]{$langle Xrangle-Delta$};
              draw[latex-] (MR) to[out=-90,in=135] ++ (0.6,-0.6) node[below right,inner
              sep=1pt]{$langle Xrangle+Delta$};
              draw[latex-] (MM) --++ (0,-0.6) node[below,inner
              sep=1pt]{$langle Xrangle$};
              end{tikzpicture}
              end{document}


              enter image description here






              share|improve this answer














              This is only a partial answer since it is not clear to me what an asymmetric Gauss curve precisely is. This is more to discuss how to set this up in principle. So I am only going to discuss how to plot a deformed Gauss curve.



              To this end, I'd like to convince you to use declare function rather than the definition you use. In the example below, I am going to use



              declare function={Gauss(x,y,z,u)=1/(z*sqrt(2*pi))*exp(-((x-y+u*(x-y)*sign(x-y))^2)/(2*z^2));


              Here Gauss reduces to an ordinary Gaussian for u=0, where x is just the variable, y defines the location of the maximum and z the width. If you turn on a nontrivial u, the Gaussian will get deformed.



              documentclass[border=5mm]{standalone}
              usepackage{pgfplots}
              pgfplotsset{height=4cm,width=8cm,compat=1.16}
              begin{document}

              begin{tikzpicture}[font=sffamily,
              declare function={Gauss(x,y,z,u)=1/(z*sqrt(2*pi))*exp(-((x-y+u*(x-y)*sign(x-y))^2)/(2*z^2));},
              every pin edge/.style={latex-,line width=1.5pt},
              every pin/.style={fill=yellow!50,rectangle,rounded corners=3pt,font=small}]
              begin{axis}[
              every axis plot post/.append style={
              mark=none,samples=101},
              clip=false,
              axis y line=none,
              axis x line*=bottom,
              ymin=0,
              xtick=empty,]
              addplot[line width=1.5pt,blue,domain=-1:3] {Gauss(x,0,0.6,-0.4)};
              draw[line width=1.5pt,dashed, black] (0,0) -- (0,{Gauss(0,0,0.6,-0.4)});
              %node[pin=270:{$X=M_e=M_o$}] at (axis cs:0,0) {};
              draw[line width=1.5pt,dashed, red] (0.6,0) -- (0.6,{Gauss(0.6,0,0.6,-0.4)});
              draw[line width=1.5pt,dashed, red] (-0.6,0) -- (-0.6,{Gauss(-0.6,0,0.6,-0.4)});
              path (-0.6,0) coordinate (ML) (0.6,0) coordinate (MR) (0,0) coordinate (MM);
              end{axis}
              draw[latex-] (ML) to[out=-90,in=45] ++ (-0.6,-0.6) node[below left,inner
              sep=1pt]{$langle Xrangle-Delta$};
              draw[latex-] (MR) to[out=-90,in=135] ++ (0.6,-0.6) node[below right,inner
              sep=1pt]{$langle Xrangle+Delta$};
              draw[latex-] (MM) --++ (0,-0.6) node[below,inner
              sep=1pt]{$langle Xrangle$};
              end{tikzpicture}
              end{document}


              enter image description here







              share|improve this answer














              share|improve this answer



              share|improve this answer








              edited Nov 26 at 4:52

























              answered Nov 26 at 3:06









              marmot

              82.3k493176




              82.3k493176












              • @ marmot They are two separate graphs, one that leans to the right and the other to the left, known as negative skew and positive skew. en.wikipedia.org/wiki/Skewness
                – Samuel Diaz
                Nov 26 at 3:25












              • @SamuelDiaz Thanks for the link! But as far as I can see it does not really give you a unique parametrization of these deformed Gaussians, does it?
                – marmot
                Nov 26 at 3:59










              • Correct, you have to keep in mind that you meet that average < average < mode or average > median > fashion.
                – Samuel Diaz
                Nov 26 at 4:09












              • @SamuelDiaz I added a possible way how you could use this.
                – marmot
                Nov 26 at 4:33


















              • @ marmot They are two separate graphs, one that leans to the right and the other to the left, known as negative skew and positive skew. en.wikipedia.org/wiki/Skewness
                – Samuel Diaz
                Nov 26 at 3:25












              • @SamuelDiaz Thanks for the link! But as far as I can see it does not really give you a unique parametrization of these deformed Gaussians, does it?
                – marmot
                Nov 26 at 3:59










              • Correct, you have to keep in mind that you meet that average < average < mode or average > median > fashion.
                – Samuel Diaz
                Nov 26 at 4:09












              • @SamuelDiaz I added a possible way how you could use this.
                – marmot
                Nov 26 at 4:33
















              @ marmot They are two separate graphs, one that leans to the right and the other to the left, known as negative skew and positive skew. en.wikipedia.org/wiki/Skewness
              – Samuel Diaz
              Nov 26 at 3:25






              @ marmot They are two separate graphs, one that leans to the right and the other to the left, known as negative skew and positive skew. en.wikipedia.org/wiki/Skewness
              – Samuel Diaz
              Nov 26 at 3:25














              @SamuelDiaz Thanks for the link! But as far as I can see it does not really give you a unique parametrization of these deformed Gaussians, does it?
              – marmot
              Nov 26 at 3:59




              @SamuelDiaz Thanks for the link! But as far as I can see it does not really give you a unique parametrization of these deformed Gaussians, does it?
              – marmot
              Nov 26 at 3:59












              Correct, you have to keep in mind that you meet that average < average < mode or average > median > fashion.
              – Samuel Diaz
              Nov 26 at 4:09






              Correct, you have to keep in mind that you meet that average < average < mode or average > median > fashion.
              – Samuel Diaz
              Nov 26 at 4:09














              @SamuelDiaz I added a possible way how you could use this.
              – marmot
              Nov 26 at 4:33




              @SamuelDiaz I added a possible way how you could use this.
              – marmot
              Nov 26 at 4:33










              up vote
              4
              down vote













              Probably slightly overkill and certainly not efficient, but you could try an approximate skew normal (central tendencies are omitted below):



              documentclass[border=5, tikz]{standalone}
              usetikzlibrary{math}
              usepackage{pgfplots}
              pgfplotsset{compat=1.14}
              tikzmath{%
              function h1(x, lx) { return (9*lx + 3*((lx)^2) + ((lx)^3)/3 + 9); };
              function h2(x, lx) { return (3*lx - ((lx)^3)/3 + 4); };
              function h3(x, lx) { return (9*lx - 3*((lx)^2) + ((lx)^3)/3 + 7); };
              function skewnorm(x, l) {
              x = (l < 0) ? -x : x;
              l = abs(l);
              e = exp(-(x^2)/2);
              return (l == 0) ? 1 / sqrt(2 * pi) * e: (
              (x < -3/l) ? 0 : (
              (x < -1/l) ? e / (8 * sqrt(2 * pi)) * h1(x, x*l) : (
              (x < 1/l) ? e / (4 * sqrt(2 * pi)) * h2(x, x*l) : (
              (x < 3/l) ? e / (8 * sqrt(2 * pi)) * h3(x, x*l) : (
              sqrt(2/pi) * e)))));
              };
              }
              begin{document}
              begin{tikzpicture}[line join=round, line cap=round]
              begin{axis}[
              width=4in, height=2in,
              every axis plot post/.append style={
              mark=none, domain=-3.5:3.5, samples=200, very thick
              },
              clip=false,
              axis y line=none,
              axis x line*=bottom,
              ymin=0, ymax=0.75,
              xtick=empty,]
              addplot[red] {skewnorm(x, -4)};
              addplot[green] {skewnorm(x, -2)};
              addplot[gray] {skewnorm(x, 0)};
              addplot[blue] {skewnorm(x, 2)};
              addplot[orange] {skewnorm(x, 4)};
              legend{$lambda=-4$,$lambda=-2$,$lambda=0$,$lambda=2$,$lambda=4$}
              end{axis}
              end{tikzpicture}
              end{document}


              enter image description here






              share|improve this answer

























                up vote
                4
                down vote













                Probably slightly overkill and certainly not efficient, but you could try an approximate skew normal (central tendencies are omitted below):



                documentclass[border=5, tikz]{standalone}
                usetikzlibrary{math}
                usepackage{pgfplots}
                pgfplotsset{compat=1.14}
                tikzmath{%
                function h1(x, lx) { return (9*lx + 3*((lx)^2) + ((lx)^3)/3 + 9); };
                function h2(x, lx) { return (3*lx - ((lx)^3)/3 + 4); };
                function h3(x, lx) { return (9*lx - 3*((lx)^2) + ((lx)^3)/3 + 7); };
                function skewnorm(x, l) {
                x = (l < 0) ? -x : x;
                l = abs(l);
                e = exp(-(x^2)/2);
                return (l == 0) ? 1 / sqrt(2 * pi) * e: (
                (x < -3/l) ? 0 : (
                (x < -1/l) ? e / (8 * sqrt(2 * pi)) * h1(x, x*l) : (
                (x < 1/l) ? e / (4 * sqrt(2 * pi)) * h2(x, x*l) : (
                (x < 3/l) ? e / (8 * sqrt(2 * pi)) * h3(x, x*l) : (
                sqrt(2/pi) * e)))));
                };
                }
                begin{document}
                begin{tikzpicture}[line join=round, line cap=round]
                begin{axis}[
                width=4in, height=2in,
                every axis plot post/.append style={
                mark=none, domain=-3.5:3.5, samples=200, very thick
                },
                clip=false,
                axis y line=none,
                axis x line*=bottom,
                ymin=0, ymax=0.75,
                xtick=empty,]
                addplot[red] {skewnorm(x, -4)};
                addplot[green] {skewnorm(x, -2)};
                addplot[gray] {skewnorm(x, 0)};
                addplot[blue] {skewnorm(x, 2)};
                addplot[orange] {skewnorm(x, 4)};
                legend{$lambda=-4$,$lambda=-2$,$lambda=0$,$lambda=2$,$lambda=4$}
                end{axis}
                end{tikzpicture}
                end{document}


                enter image description here






                share|improve this answer























                  up vote
                  4
                  down vote










                  up vote
                  4
                  down vote









                  Probably slightly overkill and certainly not efficient, but you could try an approximate skew normal (central tendencies are omitted below):



                  documentclass[border=5, tikz]{standalone}
                  usetikzlibrary{math}
                  usepackage{pgfplots}
                  pgfplotsset{compat=1.14}
                  tikzmath{%
                  function h1(x, lx) { return (9*lx + 3*((lx)^2) + ((lx)^3)/3 + 9); };
                  function h2(x, lx) { return (3*lx - ((lx)^3)/3 + 4); };
                  function h3(x, lx) { return (9*lx - 3*((lx)^2) + ((lx)^3)/3 + 7); };
                  function skewnorm(x, l) {
                  x = (l < 0) ? -x : x;
                  l = abs(l);
                  e = exp(-(x^2)/2);
                  return (l == 0) ? 1 / sqrt(2 * pi) * e: (
                  (x < -3/l) ? 0 : (
                  (x < -1/l) ? e / (8 * sqrt(2 * pi)) * h1(x, x*l) : (
                  (x < 1/l) ? e / (4 * sqrt(2 * pi)) * h2(x, x*l) : (
                  (x < 3/l) ? e / (8 * sqrt(2 * pi)) * h3(x, x*l) : (
                  sqrt(2/pi) * e)))));
                  };
                  }
                  begin{document}
                  begin{tikzpicture}[line join=round, line cap=round]
                  begin{axis}[
                  width=4in, height=2in,
                  every axis plot post/.append style={
                  mark=none, domain=-3.5:3.5, samples=200, very thick
                  },
                  clip=false,
                  axis y line=none,
                  axis x line*=bottom,
                  ymin=0, ymax=0.75,
                  xtick=empty,]
                  addplot[red] {skewnorm(x, -4)};
                  addplot[green] {skewnorm(x, -2)};
                  addplot[gray] {skewnorm(x, 0)};
                  addplot[blue] {skewnorm(x, 2)};
                  addplot[orange] {skewnorm(x, 4)};
                  legend{$lambda=-4$,$lambda=-2$,$lambda=0$,$lambda=2$,$lambda=4$}
                  end{axis}
                  end{tikzpicture}
                  end{document}


                  enter image description here






                  share|improve this answer












                  Probably slightly overkill and certainly not efficient, but you could try an approximate skew normal (central tendencies are omitted below):



                  documentclass[border=5, tikz]{standalone}
                  usetikzlibrary{math}
                  usepackage{pgfplots}
                  pgfplotsset{compat=1.14}
                  tikzmath{%
                  function h1(x, lx) { return (9*lx + 3*((lx)^2) + ((lx)^3)/3 + 9); };
                  function h2(x, lx) { return (3*lx - ((lx)^3)/3 + 4); };
                  function h3(x, lx) { return (9*lx - 3*((lx)^2) + ((lx)^3)/3 + 7); };
                  function skewnorm(x, l) {
                  x = (l < 0) ? -x : x;
                  l = abs(l);
                  e = exp(-(x^2)/2);
                  return (l == 0) ? 1 / sqrt(2 * pi) * e: (
                  (x < -3/l) ? 0 : (
                  (x < -1/l) ? e / (8 * sqrt(2 * pi)) * h1(x, x*l) : (
                  (x < 1/l) ? e / (4 * sqrt(2 * pi)) * h2(x, x*l) : (
                  (x < 3/l) ? e / (8 * sqrt(2 * pi)) * h3(x, x*l) : (
                  sqrt(2/pi) * e)))));
                  };
                  }
                  begin{document}
                  begin{tikzpicture}[line join=round, line cap=round]
                  begin{axis}[
                  width=4in, height=2in,
                  every axis plot post/.append style={
                  mark=none, domain=-3.5:3.5, samples=200, very thick
                  },
                  clip=false,
                  axis y line=none,
                  axis x line*=bottom,
                  ymin=0, ymax=0.75,
                  xtick=empty,]
                  addplot[red] {skewnorm(x, -4)};
                  addplot[green] {skewnorm(x, -2)};
                  addplot[gray] {skewnorm(x, 0)};
                  addplot[blue] {skewnorm(x, 2)};
                  addplot[orange] {skewnorm(x, 4)};
                  legend{$lambda=-4$,$lambda=-2$,$lambda=0$,$lambda=2$,$lambda=4$}
                  end{axis}
                  end{tikzpicture}
                  end{document}


                  enter image description here







                  share|improve this answer












                  share|improve this answer



                  share|improve this answer










                  answered Nov 26 at 16:02









                  Mark Wibrow

                  61.4k4111176




                  61.4k4111176






















                      up vote
                      2
                      down vote













                      Another possible way (apart from @marmot's answer) is to plot the skewed distribution function is to exploit the chi-square distribution.



                      For instance:



                      documentclass{standalone}
                      usepackage{pgfplots}
                      %https://en.wikipedia.org/wiki/Chi-squared_distribution
                      %https://tex.stackexchange.com/questions/120441/plot-the-probability-density-function-of-the-gamma-distribution?rq=1
                      % the second link gives the numerical approximation of gamma function
                      begin{document}
                      begin{tikzpicture}[
                      declare function={gamma(z)=
                      (2.506628274631*sqrt(1/z) + 0.20888568*(1/z)^(1.5) + 0.00870357*(1/z)^(2.5) - (174.2106599*(1/z)^(3.5))/25920 - (715.6423511*(1/z)^(4.5))/1244160)*exp((-ln(1/z)-1)*z);},
                      declare function={chipdf(x,k) = x^(k/2-1)*exp(-x/2) / (2^(k/2)*gamma(k));}
                      ]

                      begin{axis}[
                      axis lines=left,
                      enlargelimits=upper,
                      ]
                      addplot [smooth, domain=0:20, blue] {chipdf(x,2)};
                      addplot [smooth, domain=0:20, green] {chipdf(x,3)};
                      addplot [smooth, domain=0:20, black] {chipdf(x,4)};
                      addplot [smooth, domain=0:20, cyan] {chipdf(x,5)};
                      addplot [smooth, domain=0:20, magenta] {chipdf(x,6)};
                      end{axis}
                      end{tikzpicture}
                      end{document}


                      which will give you:



                      enter image description here



                      Using this, you can include your mean-median-mode as lines in the plot.






                      share|improve this answer

























                        up vote
                        2
                        down vote













                        Another possible way (apart from @marmot's answer) is to plot the skewed distribution function is to exploit the chi-square distribution.



                        For instance:



                        documentclass{standalone}
                        usepackage{pgfplots}
                        %https://en.wikipedia.org/wiki/Chi-squared_distribution
                        %https://tex.stackexchange.com/questions/120441/plot-the-probability-density-function-of-the-gamma-distribution?rq=1
                        % the second link gives the numerical approximation of gamma function
                        begin{document}
                        begin{tikzpicture}[
                        declare function={gamma(z)=
                        (2.506628274631*sqrt(1/z) + 0.20888568*(1/z)^(1.5) + 0.00870357*(1/z)^(2.5) - (174.2106599*(1/z)^(3.5))/25920 - (715.6423511*(1/z)^(4.5))/1244160)*exp((-ln(1/z)-1)*z);},
                        declare function={chipdf(x,k) = x^(k/2-1)*exp(-x/2) / (2^(k/2)*gamma(k));}
                        ]

                        begin{axis}[
                        axis lines=left,
                        enlargelimits=upper,
                        ]
                        addplot [smooth, domain=0:20, blue] {chipdf(x,2)};
                        addplot [smooth, domain=0:20, green] {chipdf(x,3)};
                        addplot [smooth, domain=0:20, black] {chipdf(x,4)};
                        addplot [smooth, domain=0:20, cyan] {chipdf(x,5)};
                        addplot [smooth, domain=0:20, magenta] {chipdf(x,6)};
                        end{axis}
                        end{tikzpicture}
                        end{document}


                        which will give you:



                        enter image description here



                        Using this, you can include your mean-median-mode as lines in the plot.






                        share|improve this answer























                          up vote
                          2
                          down vote










                          up vote
                          2
                          down vote









                          Another possible way (apart from @marmot's answer) is to plot the skewed distribution function is to exploit the chi-square distribution.



                          For instance:



                          documentclass{standalone}
                          usepackage{pgfplots}
                          %https://en.wikipedia.org/wiki/Chi-squared_distribution
                          %https://tex.stackexchange.com/questions/120441/plot-the-probability-density-function-of-the-gamma-distribution?rq=1
                          % the second link gives the numerical approximation of gamma function
                          begin{document}
                          begin{tikzpicture}[
                          declare function={gamma(z)=
                          (2.506628274631*sqrt(1/z) + 0.20888568*(1/z)^(1.5) + 0.00870357*(1/z)^(2.5) - (174.2106599*(1/z)^(3.5))/25920 - (715.6423511*(1/z)^(4.5))/1244160)*exp((-ln(1/z)-1)*z);},
                          declare function={chipdf(x,k) = x^(k/2-1)*exp(-x/2) / (2^(k/2)*gamma(k));}
                          ]

                          begin{axis}[
                          axis lines=left,
                          enlargelimits=upper,
                          ]
                          addplot [smooth, domain=0:20, blue] {chipdf(x,2)};
                          addplot [smooth, domain=0:20, green] {chipdf(x,3)};
                          addplot [smooth, domain=0:20, black] {chipdf(x,4)};
                          addplot [smooth, domain=0:20, cyan] {chipdf(x,5)};
                          addplot [smooth, domain=0:20, magenta] {chipdf(x,6)};
                          end{axis}
                          end{tikzpicture}
                          end{document}


                          which will give you:



                          enter image description here



                          Using this, you can include your mean-median-mode as lines in the plot.






                          share|improve this answer












                          Another possible way (apart from @marmot's answer) is to plot the skewed distribution function is to exploit the chi-square distribution.



                          For instance:



                          documentclass{standalone}
                          usepackage{pgfplots}
                          %https://en.wikipedia.org/wiki/Chi-squared_distribution
                          %https://tex.stackexchange.com/questions/120441/plot-the-probability-density-function-of-the-gamma-distribution?rq=1
                          % the second link gives the numerical approximation of gamma function
                          begin{document}
                          begin{tikzpicture}[
                          declare function={gamma(z)=
                          (2.506628274631*sqrt(1/z) + 0.20888568*(1/z)^(1.5) + 0.00870357*(1/z)^(2.5) - (174.2106599*(1/z)^(3.5))/25920 - (715.6423511*(1/z)^(4.5))/1244160)*exp((-ln(1/z)-1)*z);},
                          declare function={chipdf(x,k) = x^(k/2-1)*exp(-x/2) / (2^(k/2)*gamma(k));}
                          ]

                          begin{axis}[
                          axis lines=left,
                          enlargelimits=upper,
                          ]
                          addplot [smooth, domain=0:20, blue] {chipdf(x,2)};
                          addplot [smooth, domain=0:20, green] {chipdf(x,3)};
                          addplot [smooth, domain=0:20, black] {chipdf(x,4)};
                          addplot [smooth, domain=0:20, cyan] {chipdf(x,5)};
                          addplot [smooth, domain=0:20, magenta] {chipdf(x,6)};
                          end{axis}
                          end{tikzpicture}
                          end{document}


                          which will give you:



                          enter image description here



                          Using this, you can include your mean-median-mode as lines in the plot.







                          share|improve this answer












                          share|improve this answer



                          share|improve this answer










                          answered Nov 26 at 7:19









                          Raaja

                          2,0262527




                          2,0262527






























                              draft saved

                              draft discarded




















































                              Thanks for contributing an answer to TeX - LaTeX Stack Exchange!


                              • Please be sure to answer the question. Provide details and share your research!

                              But avoid



                              • Asking for help, clarification, or responding to other answers.

                              • Making statements based on opinion; back them up with references or personal experience.


                              To learn more, see our tips on writing great answers.





                              Some of your past answers have not been well-received, and you're in danger of being blocked from answering.


                              Please pay close attention to the following guidance:


                              • Please be sure to answer the question. Provide details and share your research!

                              But avoid



                              • Asking for help, clarification, or responding to other answers.

                              • Making statements based on opinion; back them up with references or personal experience.


                              To learn more, see our tips on writing great answers.




                              draft saved


                              draft discarded














                              StackExchange.ready(
                              function () {
                              StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2ftex.stackexchange.com%2fquestions%2f461758%2fasymmetric-distribution-gauss-curve%23new-answer', 'question_page');
                              }
                              );

                              Post as a guest















                              Required, but never shown





















































                              Required, but never shown














                              Required, but never shown












                              Required, but never shown







                              Required, but never shown

































                              Required, but never shown














                              Required, but never shown












                              Required, but never shown







                              Required, but never shown







                              Popular posts from this blog

                              Plaza Victoria

                              In PowerPoint, is there a keyboard shortcut for bulleted / numbered list?

                              How to put 3 figures in Latex with 2 figures side by side and 1 below these side by side images but in...