Scatter plots¶

Using stackview.scatterplot we can visualize contents of pandas DataFrames. In such a plot you can select objects and visualize the selection. This might be useful for exploring feature extraction parameter spaces.

In [1]:

import pandas as pd
import numpy as np
import stackview
import pandas as pd
from skimage.measure import regionprops_table
from skimage.io import imread
from skimage.filters import threshold_otsu
from skimage.measure import label
import matplotlib.pyplot as plt

stackview.__version__

Out[1]:

'0.7.11'

To demonstrate this, we need an image, a segmentation and a table of extracted features.

In [2]:

image = imread('data/blobs.tif')

# segment image
thresh = threshold_otsu(image)
binary_image = image > thresh
labeled_image = label(binary_image)

In [3]:

properties = regionprops_table(labeled_image, properties=['centroid', 'area', 'feret_diameter_max', 'minor_axis_length', 'major_axis_length'])

df = pd.DataFrame(properties)
df["aspect_ratio"] = df['major_axis_length'] / df['minor_axis_length']

num_objects = df.shape[0]
pre_selection = np.zeros(num_objects)
pre_selection[:int(num_objects/2)] = 1

pre_selection2 = np.zeros(num_objects)
pre_selection2[::2] = 1

df["selection"] = pre_selection

df.head(6)

Out[3]:

	centroid-0	centroid-1	area	feret_diameter_max	minor_axis_length	major_axis_length	aspect_ratio	selection
0	13.212471	19.986143	433.0	36.055513	16.819060	34.957399	2.078439	1.0
1	4.270270	62.945946	185.0	21.377558	11.803854	21.061417	1.784283	1.0
2	12.568389	108.329787	658.0	32.449961	28.278264	30.212552	1.068402	1.0
3	9.806452	154.520737	434.0	26.925824	23.064079	24.535398	1.063793	1.0
4	13.545073	246.809224	477.0	31.384710	19.833058	31.162612	1.571246	1.0
5	15.757895	197.210526	285.0	21.840330	17.770536	20.506730	1.153974	1.0

Drawing an interactive scatter plot¶

Next we draw a scatter plot of area versus Feret's diameter. You can select the axes which are plotted and the plot will change. You can also use a lasso-tool to select data points.

In [4]:

stackview.scatterplot(df, 'area', 'feret_diameter_max', "selection", figsize=(5,4))

Out[4]:

HBox(children=(VBox(children=(VBox(children=(HBox(children=(Label(value='Axes '), Dropdown(index=2, layout=Lay…

Next we take the selection from the dataframe and visualize this as an image.

In [5]:

selection = df["selection"].tolist()

selected_image = np.take(np.asarray([-1] + list(selection)) * 1 + 1, labeled_image).astype(np.uint32)

stackview.insight(selected_image)

Out[5]:

shape	(254, 256)
dtype	uint32
size	254.0 kB
min	0
max	2

Interaction¶

Using some more involved code we can also draw the image and the scatter plot side-by-side and make them interact. You can select data points in the plot on the right and the visualization on the left will be updated accordingly.

In [6]:

stackview.clusterplot(image=image,
                     labels=labeled_image,
                     df=df,
                     column_x="area",
                     column_y="aspect_ratio", 
                     zoom_factor=1.6,
                     alpha=0.7)

Out[6]:

VBox(children=(HBox(children=(VBox(children=(HBox(children=(VBox(children=(ImageWidget(height=406, width=409),…

Every time the user selects different data points, the selection in our dataframe is update

In [7]:

df["selection"]

Out[7]:

0     1.0
1     1.0
2     1.0
3     1.0
4     1.0
     ... 
59    0.0
60    0.0
61    0.0
62    0.0
63    0.0
Name: selection, Length: 64, dtype: float64

Custom widgets¶

You can also build such a side-by-side view with two plots yourself:

In [10]:

import ipywidgets as widgets
from ipywidgets import HBox

# Program functions that are executed when the selection is changed.
def update2(e=None):
    widget2.update()

def update1(e=None):
    widget1.update()
    
widget1 = stackview.scatterplot(df, column_x="centroid-0", column_y="centroid-1", selection_changed_callback=update2, markersize=50)
widget2 = stackview.scatterplot(df, column_x="area", column_y="aspect_ratio", selection_changed_callback=update1)

# Arrange the widgets side by side using HBox
HBox([widget1, widget2])

Out[10]:

HBox(children=(HBox(children=(VBox(children=(VBox(children=(HBox(children=(Label(value='Axes '), Dropdown(layo…

In [ ]: