unexpected histogram bin size behavior

If the maximum number in the interval is a multiple of the bin size, then this number is lost.
Here is a code example:

var x = [1.0, 9.1, 10.0, 17.8, 18.5, 20.0];

var trace = {
    x: x,
    type: 'histogram',
    autobinx: false,
    xbins: { 
      end: 20.0, 
      size: 10.0,
      start: 0
    }
};
var data = [trace];
Plotly.newPlot('myDiv', data);

This is the result of using the notation [a; b) when creating groups. This behavior creates a problem, for example, when you want to dynamically adjust the bin size depending on the step. Here is an example:

var data = [1.0, 9.1, 10.0, 17.8, 18.5, 20.0];

step = 1.0; // any value from the interval: [0; 1]

xbins_start = data.reduce((x, y) => Math.min(x, y), 0.0);
xbins_end   = data.reduce((x, y) => Math.max(x, y), 0.0);
xbins_size  = (step * (xbins_end - xbins_start));

var trace = {
    x: data,
    type: 'histogram',
    autobinx: false,
    xbins: { 
      start: xbins_start,
      end: xbins_end,
      size: xbins_size
    }
};

Plotly.newPlot('myDiv', [trace]);

By setting the step size to 1, I hope to get the only one group containing all the data values (the height of the bin should be 6), instead I get a height of 5 with the loss of the maximum value.

Commencez par la reproduction JavaScript dans l’issue et suivez le chemin de binning de l’histogramme pour les valeurs explicites start, end et size de xbins. Vérifiez le comportement aux limites avec les données fournies et confirmez qu’une valeur égale à end est comptée dans le bin résultant.