quiver: nan issue when u or v entries have a 0 #279
Description
Bug category
- bug - compilation error
- bug - compilation warning
- bug - runtime error
- bug - runtime warning
- bug - logic error
Describe the bug
When an entry for u or v is all zeros, the scaling results in u_scaled/v_scaled to have nans due to a division by u_max, v_max.
Steps to Reproduce
#include <cmath>
#include <matplot/matplot.h>
int main() {
using namespace matplot;
// Produces 10 entries
auto x = iota(0, 2.0 / 20, 2);
auto y = iota(0, 2.0 / 20, 2);
// Produces 10 entries
std::vector<double> theta;
for (int t = 0; t < x.size(); t++) {
theta.push_back(0 + 2 * M_PI / x.size() * t);
}
auto u =
transform(theta, [](double theta) { return 1; });
auto v =
transform(theta, [](double theta) { return 0; });
quiver(x, y, u, v);
show();
return 0;
}Output
Details
Empty plot with the following console output:Press ENTER to continue... line 21: warning: Skipping data file with no valid points
multiplot> plot '-' with vectors linecolor rgb "0x000071BC" dashtype 1 linewidth 0.5 linetype -1
^
line 21: x range is invalid
line 42: warning: Skipping data file with no valid points
multiplot> plot '-' with vectors linecolor rgb "0x000071BC" dashtype 1 linewidth 0.5 linetype -1
^
line 42: x range is invalid
line 42: warning: refresh not possible and replot is disabled
line 42: warning: refresh not possible and replot is disabled
line 42: warning: refresh not possible and replot is disabled
line 42: warning: refresh not possible and replot is disabledPlatform
- cross-platform issue - linux
- cross-platform issue - windows
- cross-platform issue - macos
Environment Details:
- OS: Archlinux
- OS Version: Rolling
- Compiler:
gcc - Compiler version:
gcc-12.1.0
Additional context
Related discussion, which prompted me in discovering this bug: #278