1. lets-plot
  2. docs
  3. dev
  4. notebooks
In [1]:
from lets_plot import *
from lets_plot.bistro import *
import pandas as pd
LetsPlot.setup_html()
In [2]:
df = {
    'x': [1],
    'y': [1],
    's': [1]
}

Without scale

In [3]:
ggplot(df, aes('x','y')) + geom_point() + ggtb()
Out[3]:

size_zoomin parameter

The size_zoomin parameter accepts an integer value:

  • 0 — zoom in disabled (default)
  • -1 — unlimited zoom in
  • Any other positive number — maximum zoom-in limit. For example, 2 means the geometry can be scaled up to 2×.

size_zoomin=0 zoom in disabled

In [4]:
ggplot(df, aes('x','y')) + geom_point() + ggtb(size_zoomin=0)
Out[4]:

size_zoomin=-1 unlimited zoom in

In [5]:
ggplot(df, aes('x','y')) + geom_point() + ggtb(size_zoomin=-1)
Out[5]:

size_zoomin=3 zoom is capped at three times.

In [6]:
ggplot(df, aes('x','y')) + geom_point() + ggtb(size_zoomin=3)
Out[6]:

Illegal size_zoomin value

In [7]:
ggplot(df, aes('x','y')) + geom_point() + ggtb(size_zoomin=-3)
Out[7]:

size_basis parameter

The size_basis parameter accepts a string value: x, y, min, or max (default: max). It defines which axis is used to calculate the scaling factor.

  • x and y specify the corresponding axis.
  • min uses the smaller scaling factor.
  • max uses the larger scaling factor.

size_basis="x"

Try using rubber-band zoom along the x axis only (and then along the y axis) to see the effect.

In [8]:
ggplot(df, aes('x','y')) + geom_point(size=10) + ggtb(size_basis="x", size_zoomin=-1)
Out[8]:

size_basis="y"

In [9]:
ggplot(df, aes('x','y')) + geom_point(size=10) + ggtb(size_basis="y", size_zoomin=-1)
Out[9]:

size_basis="min"

In [10]:
ggplot(df, aes('x','y')) + geom_point(size=10) + ggtb(size_basis="min", size_zoomin=-1)
Out[10]:

size_basis="max" default

In [11]:
ggplot(df, aes('x','y')) + geom_point(size=10) + ggtb(size_basis="max", size_zoomin=-1)
Out[11]:

Support for various size types

Default size

In [12]:
ggplot(df, aes('x','y')) + geom_point() + ggtb(size_zoomin=-1)
Out[12]:

Constant size

In [13]:
ggplot(df, aes('x','y')) + geom_point(size=80) + ggtb(size_zoomin=-1)
Out[13]:

Size from Aes

In [14]:
ggplot(df, aes('x','y')) + geom_point(aes(size='s')) + ggtb(size_zoomin=-1)
Out[14]:

Constant size with size_unit

Scaling from ggtb() is not applied to layers that use size_unit. However, it is still applied to the other layers.

In [15]:
ggplot(df, aes('x','y')) + \
    geom_point(size=0.1, size_unit = 'x') + \
    geom_point(size=2, color='red') + \
    ggtb(size_zoomin=-1)
Out[15]:

Support gggrid()

ggtb() can only be added to the root gggrid(). All plots inside will use the same scaling settings. Adding ggtb() to plots inside gggrid() is not allowed.

In [16]:
gggrid(
    [
        ggplot(df, aes('x','y')) + geom_point(size=10),
        gggrid([
            ggplot(df, aes('x','y')) + geom_point(size=10),
            ggplot(df, aes('x','y')) + geom_point(size=20)
        ])
    ]
) + ggtb(size_zoomin=-1, size_basis="x")
Out[16]:

Support facet_wrap()

In [17]:
mpg = pd.read_csv('https://raw.githubusercontent.com/JetBrains/lets-plot-docs/master/data/mpg2.csv')
In [18]:
p = (
  ggplot(mpg, 
         aes('engine horsepower', 'miles per gallon', 
             color=as_discrete('origin of car',order=-1)))
  + geom_point()
  + theme_grey() + ggtitle('Efficiency vs Engine Horsepower')
)
p
Out[18]:
In [19]:
p + facet_wrap(['origin of car', 'number of cylinders'], order=[-1, 1], ncol=3) + ggtb(size_zoomin=3)
Out[19]:
In [ ]: