How to update histogram nbins with buttons in plotly (python)?-CodePudding

I am having trouble updating the number of bins when updating the plot with buttons. This is an examplary short dataframe. In the full version, there are a lot more rows so there is a point in using the initial nbins = 100. However, I would like to change the number of bins for each column when updating the histogram.

dataset = pd.DataFrame(
    {'age': [19, 18, 28, 33, 32],
    'bmi': [27.9, 33.77, 33.0, 22.705, 28.88],
    'children': [0, 1, 3, 0, 0]}
)

fig = px.histogram(x = dataset['age'], nbins = 100)
fig.update_layout(xaxis_title = 'age')

buttons = []
for column in dataset.columns:
    buttons.append(
        dict(
            args = [
                {'x': [dataset[column]]},
                {'xaxis.title': column},
                
                # __________________________________
                # I TRIED THIS since fig.layout.figure.data[0].nbinsx = 100
                # {'figure.data[0].nbinsx': 5}
                # __________________________________

            ],
            label = column,
            method = 'update',
        )
    )

fig.update_layout(
    updatemenus = [
        dict(type = 'buttons', buttons = buttons,
             direction = 'right', x=1, y=1.15)
    ],
    title_text = 'Histograms'
)
fig.show()

This is how the histogram looks with the button options available.

**When I change the column that the histogram is constructed for, the number of bins do not change. How do I fix this? I tried to **

THIS IS THE IMAGE OF THE HISTOGRAM! I HAVE YET TO EARN REPUTATION POINTS FOR IMAGE EMBEDDING TO BE POSSIBLE.

enter image description here

CodePudding user response：

you have not defined number of bins you want per column. Have used col_bins = {c: int(dataset[c].max()-dataset[c].min()) for c in dataset.columns} to generate values from your data
from a style perspective I use list and dict comprehensions to build updatemenus structures
your code is close, key part is understanding args of update method. It's a list where first element is dict to update traces and second element is a dict to update layout. There are two updates to traces: x and nbinsx so they are including in one dict

import pandas as pd
import plotly.express as px

dataset = pd.DataFrame(
    {
        "age": [19, 18, 28, 33, 32],
        "bmi": [27.9, 33.77, 33.0, 22.705, 28.88],
        "children": [0, 1, 3, 0, 0],
    }
)
col_bins = {c: int(dataset[c].max()-dataset[c].min()) for c in dataset.columns}

fig = px.histogram(x=dataset["age"], nbins=100)
fig.update_layout(
    updatemenus=[
        {
            "buttons": [
                {
                    "label": c,
                    "method": "update",
                    "args": [{"x": [dataset[c]], "nbinsx":bins}, {"xaxis.title":c}],
                }
                for c, bins in col_bins.items()
            ],
            "direction": "right",
            "type":"buttons",
            "x":1,
            "y":1.15
        }
    ],
    xaxis_title="age",
    title_text = 'Histograms'
)