mdadash.backend.analyses.custom_code

Custom user-defined code

Classes

CustomCode()

Custom Code

class mdadash.backend.analyses.custom_code.CustomCode

Bases: WidgetBase

Custom Code

This widget allows custom user-defined code to run during widget execution. The run frequency can be every-frame or batch similar to other widgets. User-defined code can be split into two parts - Setup code and Execute code as shown below:

The code under Setup code is executed only once at the time of widget creation. Examples of code that goes here would be common functions, class definitions, initializations etc.

The code under Execute code is executed as per the chosen run frequency. The outputs of this code will show up as the widget outputs. The outputs can include both text and images. Any errors, including syntax errors and code errors will show up in the output as well.

The global variable u points to the current MDAnalysis Universe and can be used directly in the code. There is support for code complete and inspect as shown below:

Examples

Example 1: Box volume

To print the current box volume, the following can be used in the Execute code:

print(f"Box Volume = {u.trajectory.ts.volume:.2f} ų")

Example 2: Kinetic Energy of an AtomGroup

To print and display the Kinetic Energy of an AtomGroup a simple KE function and plot can be defined in the Setup code as follows (executes only once):

import numpy as np
from collections import deque
import matplotlib.pyplot as plt

def ke(ag):
    return 0.5 * np.sum(ag.masses[:, np.newaxis] * (ag.velocities**2))

class SimplePlot:
    def __init__(self, universe, max_values=100):
        self.u = universe
        self.times = deque(maxlen=max_values)
        self.values = deque(maxlen=max_values)

    def show(self, name, values):
        self.values.append(values)
        self.times.append(self.u.trajectory.ts.data["time"])
        plt.plot(self.times, self.values)
        plt.ylabel("Values")
        plt.xlabel("Time (ps)")
        plt.title(name)
        plt.grid(True)
        plt.show()

plt1 = SimplePlot(u)

The above function and plot can be invoked in the Execute code section as follows:

ke1 = ke(u.select_atoms("protein"))
print(f"Protein KE = {ke1:.2f}")
plt1.show("KE (protein)", ke1)

Here is an example of the widget output when the widget is running:

Note

Variables created in custom code blocks are common across the dashboard kernel and hence must be chosen to not overwrite each other.

description = 'Custom user-defined code'

name = 'Custom Code'

on_post_connect()

on_post_connect handler

run_batch()

batch run handler

run_every_frame()

every-frame run handler