Ajouter des annotations

Visualisation de données interactive avec Bokeh

George Boorman

Core Curriculum Manager, DataCamp

Annotation de boîte

annotation de boîte

Ajouter des annotations de boîte

from bokeh.models import BoxAnnotation
fig = figure(x_axis_label="Rebounds", y_axis_label="Points", title="Points vs. Rebounds")
fig.circle(x="rebounds", y="points", source=source)

low_box = BoxAnnotation(top=10, fill_color="red", fill_alpha=0.3)

mid_box = BoxAnnotation(bottom=10, top=20, fill_color="orange", fill_alpha=0.3)

high_box = BoxAnnotation(bottom=20, fill_color="green", fill_alpha=0.3)

fig.add_layout(low_box)
fig.add_layout(mid_box)
fig.add_layout(high_box)

fig.title.align = "center"
output_file("color_annotated_plot.html")
show(fig)

Annotation polygonale

tracé avec polygone

Ajouter une annotation polygonale

import datetime as dt
from bokeh.models import PolyAnnotation
fig = figure(title="Nvidia Stock Price", x_axis_label="Date", y_axis_label="Price")
fig.line(x="date", y="close", source=source)
fig.xaxis[0].formatter = DatetimeTickFormatter(months="%b %Y")
fig.yaxis[0].formatter = NumeralTickFormatter(format="$0")

start_date = dt.datetime(2017, 5, 9)
end_date = dt.datetime(2017, 6, 12)

start_float = start_date.timestamp() * 1000
end_float = end_date.timestamp() * 1000

start_data = nvidia.loc[nvidia['date'] == start_date]['close'].values[0]
end_data = nvidia.loc[nvidia['date'] == end_date]['close'].values[0]

polygon = PolyAnnotation(fill_color="green", fill_alpha=0.4,

                         xs = [start_float, start_float, end_float, end_float],

                         ys = [start_data-8, start_data+28, end_data+20, end_data-10])

fig.add_layout(polygon)
output_file(filename="nvidia_polygon_annotation.html")
show(fig)

Passons à la pratique !

Visualisation de données interactive avec Bokeh