Frame Plotting

This provides a simple high-level interface of plotting a frame.

To start

f = Frame2D.from_image("path/to/file.png")
fpl = f.plot()

This gives you a Frame2DPlot object to call plotting functions with.

If you only want to plot a few indexes, you can call it with arguments

from sklearn.preprocessing import robust_scale,

f = Frame2D.from_image("path/to/file.png")
fpl = f.plot([0, 1])

If the file is an RGB image, it’ll just plot with R and G only.

There are multiple ways to generate the plots, listed in the Module Info.

Setting Subplot Shape

By default, if Frame2DPlot.subplot_shape is None, it’ll find the best shape to fit the plot.

However, if you want to set a fixed shape, you can do so

ROWS = 3
COLS = 2
fpl.subplot_shape = (ROWS, COLS)

Setting Titles

By default, if Frame2DPlot.titles is None, it’ll generate titles of Index 0, Index 1, … and so on.

If you do set the title, take note that the number of titles must match the number of channels, else an AssertionError will be thrown.

Example

Here, we load in an image on 25% scale then plot the images with 22 rows and 1 column.

SCALE = 0.25
f = Frame2D.from_image("path/to/file.png", scale=SCALE)
fc = f.get_all_chns()

fpl = fc.plot()
ROWS = 22
COLS = 1
PLT_SCALE = 1.1
fpl.subplot_shape = (ROWS, COLS)
fpl.image(scale=PLT_SCALE)

plt.show()

Module Info

class frmodel.base.D2.frame._frame_plot.Frame2DPlot(f: Frame2D, subplot_shape: tuple = None, titles: list = None)

Bases: object

_create_grid(scale: float = 1.0)

Facilitates in create a subplot grid for plotting functions.

Parameters

scale – Scale of the plot

Returns

An Axes reference generator

hist(scale=1, bins=50)

For each index, create a separate subplot hist.

Parameters

  • scale – Scale of the subplots

  • bins – Number of bins to pass into hist

Returns

A plt.Figure

image(scale: float = 1, colormap: str = 'magma')

For each index, create a separate subplot imshow.

Parameters

  • scale – Scale of the subplots

  • colormap – The cmap of imshow. See plt.imshow for available cmaps.

Returns

A plt.Figure

kde(scale=1, smoothing=0.5)

For each index, create a separate subplot hist.

Note: smoothing may not work on some versions of seaborn.

Parameters

  • scale – Scale of the subplots

  • smoothing – The amount of smoothing to apply to the KDE

Returns

A plt.Figure

scatter3d(chn: frmodel.base.consts.CONSTS.CHN, colored: bool = False, z_scale: int = 1, point_size: float = 7, sample_size: int = 10000, colorscale=['#440154', '#482878', '#3e4989', '#31688e', '#26828e', '#1f9e89', '#35b779', '#6ece58', '#b5de2b', '#fde725'])

Plot a single index with respect to its X and Y.

Parameters

  • chn – A single channel.

  • colored – Whether to have the point cloud coloured with RGB channels. Only works if RGB exists.

  • z_scale – The scale of the Z Axis. If lower than 1, it’ll look flatter, vice versa.

  • point_size – The size of the markers.

  • sample_size – The amount of points to sample. If None, use all points

  • colorscale – The color scale to use when plotting.

Returns

A plt.Figure

static set_browser_plotting()

Makes Plotly render on browser by changing the flag.

surface3d(chn: frmodel.base.consts.CONSTS.CHN, nan_value: float = 0)

Plots a surface 3d plot on Plotly

Parameters

  • chn – The channel to plot as height

  • nan_value – The value to replace NaNs