Graded shading of the plot, intended to improve the readability of 3D surface charts, is also applied by default to 2D contour plots, making contour bands difficult to interpret, and in this case almost unreadable.

There are only 3 contour bands, resulting in a greatly simplified plot.

The chart is inverted, with Y values increasing from top to bottom.

The Excel chart with default settings on the other hand is almost useless.

The interpolation of the contours is more jagged, because the vertical mid-side nodes were ignoredĭealing with these differences will be discussed in the next post.

The horizontal and vertical dimensions are not plotted to equal scale.

The Matplotlib results are reasonably close to the Strand7 plot, the main differences being: The resulting chart, with all default settings, is: Excel contour plot with default settings

Click on the 3rd icon from the ‘Surface’ selection (Contour).

Select the Insert tab, then the ‘Waterfall, Funnel, Stock, Surface, or Radar chart’ group.

Select the XYZ data arranged as a single range (Range I4:N8 in the data screenshot below).

The same input data shown above can be used to generate a contour plot in Excel as follows: The Matplotlib code, modified to accept input of the Z data 'float')ĭef contour_plot2(Xvals, Yvals, Z, levels, colors):Īnd the resulting output: Matplotlib contour plot To satisfy those requirements the deflections were extracted along the top and bottom edge of each plate as shown below: The contoured data is also required to be in a 2D grid. The Strand7 vertical deflection results are shown below: Strand7 Y deflection contour outputĮach plate in the Strand7 model is defined by 8 nodes (the corners and the midpoint of each side), but the Excel contour plot and the standard contourf input require a rectangular grid, with an equal number of points in each row, and for the Excel plot, equal spacing between each point.

The first FEA example is a very simple 2D plane strain analysis with 6 rectangular plates, restrained horizontally along the two vertical edges, vertically and horizontally along the base, with a downward deflection of 10 mm applied to the top left corner. X,Y,Z = mandelbrot_set5(xmin,xmax,ymin,ymax,img_width,img_height,maxiter)Ī more practical example is plotting finite element analysis results, where we will compare Matplotlib alternatives and Excel contour plots with output from the FEA program Strand7. If maxiter < maxiter1: maxiter = maxiter1 Maxiter = np.int(np.log(xrange0/xrange)*itfactor+maxiter1) The contourf function can also be used as an alternative for plotting the Mandelbrot set (although it is much slower in generating the plot than the procedures used in previous 'float')ĭef contour_mandelplot(xc,xrange,yc,yrange, maxiter1, zoomfact, itfactor, width, height, levels, colors): ‘levels’ is a single column range listing the values for each contour boundary, and ‘colors’ may be a single column range with color codes, or a three or four column range specifying the color components (see Matplotlib documentation for details).

The X and Y data is listed in single column ranges (or in this case the same range for X and Y). Input for the ‘levels’ and ‘colors’ arguments is shown in the screenshot below: # Generate the contour plot with contourf, using contour bands specified in the levels list, and colours specified in colorsĬp = ax.contourf(X, Y, Z, levels=levels, extend='both', colors= colors)įig.colorbar(cp) # Add a colorbar to a plot # If the first element of colors is a string, extract the first column and convert to a list # Calculate the Z array with Numpy functions # Use Numpy meshgrid to convert the 1D input lists to 2D grids Def contour_plot_ex1(Xvals, Yvals, levels, colors):

Contour plot in python