Useful tips

The API documentation does not give credits to some useful features find in prose. Here is a non exausthive demonstartion of some of them:

• Accessing blocks

• Plotting in Sequence

• Checking catalogs

• Custom apertures

In what follows we will make examples from a random set of images

import numpy as np
from prose import Image, Sequence

np.random.seed(42)
images = [Image(data=np.random.randn(100,100)) for i in range(5)]

Accessing blocks

When a Sequence is defined there are few ways to access its blocks. Let’s define one for example

from prose.blocks import Set, Pass, SegmentedPeaks

sequence = Sequence([
    Pass(name="named_block"),
    Set(example=True),
    SegmentedPeaks(),
])

The list of blocks is returned by

sequence.blocks

[<prose.blocks.utils.Pass at 0x29442ecd0>,
 <prose.blocks.utils.Set at 0x29442ec10>,
 <prose.blocks.detection.SegmentedPeaks at 0x105803dc0>]

but specific blocks can also be accessed by name (if they have one)

sequence.named_block

<prose.blocks.utils.Pass at 0x29442ecd0>

Plotting in Sequence

The Plot Block can be used to record any plotting function applied to an Image, like

import matplotlib.pyplot as plt

# This function plot the image, its max and an histogram of it
def show(image):
    plt.figure(figsize=(8,4))
    plt.subplot(121)
    plt.title("image data", loc="left")
    plt.imshow(image.data, origin="lower", cmap="Greys")
    plt.plot(*np.unravel_index(np.argmax(image.data), image.data.shape), "o",
             ms=10, fillstyle="none", c="w", markeredgewidth=3, label="max")
    plt.text(5, 5, image.i if hasattr(image, "i") else "--", color="white", fontsize=14)
    plt.legend()

    plt.subplot(122)
    plt.hist(image.data.flatten(), bins=100, color="k")
    plt.xlim(-5, 5)
    plt.ylim(0, 500)
    plt.title("image histogram", loc="left")
    plt.tight_layout()

# for example
show(images[0])

This can be used within a Sequence through the use of Plot to create a video/gif of it

from prose.blocks.vizualisation import PlotVideo

# we define our block
plot_block = PlotVideo(show, "static/plots.gif")

# adding to the sequence and running
sequence = Sequence([plot_block])
sequence.run(images)

RUN 100%|█████████████████████████████████████| 5/5 [00:00<00:00,  8.89images/s]

Custom apertures

Let’s start by loading the phot originated from the photometry tutorial and detect som stars on it

from prose import blocks, Observation

obs = Observation("static/example.phot")
im = obs.stack
im.show()

<AxesSubplot:>

Target is 0, we plot its PSF and define some adapted aperture(s)

import numpy as np

# defining aperture(s)
rin = 10
rout = 20
apertures = np.linspace(3, rin, 10) # multiple apertures

obs.plot_radial_psf(0, aperture=apertures[0], rin=rin, rout=rout, n=15)

This looks reasonable enough for aperture photometry to be performed (see the photometry tutorial)