|
| 1 | +Basic usage |
| 2 | +=========== |
| 3 | + |
| 4 | +The easiest way to get started with Python-LabThings is via the :mod:`labthings.quick` module, and the :class:`labthings.LabThing` builder methods. |
| 5 | + |
| 6 | +We will assume that for basic usage you already have some basic instrument control code. In our example, this is in the form of a ``PretendSpectrometer`` class, which will generate some data like your instrument control code might. Our ``PretendSpectrometer`` class has a ``data`` property which quickly returns a spectrum, an ``x_range`` property which determines the range of data we'll return, a ``magic_denoise`` property for cleaning up our signal, and a slow ``average_data(n)`` method to average ``n`` individual data measurements. |
| 7 | + |
| 8 | +Building an API from this class requires a few extra considerations. In order to tell our API what data to expect from users, we need to construct a schema for each of our interactions. This schema simply maps variable names to JSON-compatible types, and is made simple via the :mod:`labthings.fields` module. |
| 9 | + |
| 10 | +For properties, the input and output MUST be formatted the same, and so a single ``schema`` argument handles both. For actions, the input parameters and output response may be different. In this case, we can pass a ``schema`` argument to format the output, and an ``args`` argument to specify the input parameters, |
| 11 | + |
| 12 | +An example Lab Thing built from our ``PretendSpectrometer`` class, complete with schemas, might look like: |
| 13 | + |
| 14 | + |
| 15 | +.. code-block:: python |
| 16 | +
|
| 17 | + from labthings.server.quick import create_app |
| 18 | + from labthings.server import fields |
| 19 | +
|
| 20 | + from my_components import PretendSpectrometer |
| 21 | +
|
| 22 | +
|
| 23 | + # Create LabThings Flask app |
| 24 | + app, labthing = create_app( |
| 25 | + __name__, |
| 26 | + title="My PretendSpectrometer API", |
| 27 | + description="LabThing API for PretendSpectrometer", |
| 28 | + version="0.1.0" |
| 29 | + ) |
| 30 | +
|
| 31 | +
|
| 32 | + # Make some properties and actions out of our component |
| 33 | +
|
| 34 | + # Single-shot data property |
| 35 | + labthing.build_property( |
| 36 | + my_component, # Python object |
| 37 | + "data", # Objects attribute name |
| 38 | + "/data", # URL to bind the property to |
| 39 | + description="A single-shot measurement", |
| 40 | + readonly=True, |
| 41 | + schema=fields.List(fields.Number()) |
| 42 | + ) |
| 43 | +
|
| 44 | + # Magic denoise property |
| 45 | + labthing.build_property( |
| 46 | + my_component, # Python object |
| 47 | + "magic_denoise", # Objects attribute name |
| 48 | + "/denoise", # URL to bind the property to |
| 49 | + description="A magic denoise property", |
| 50 | + schema=fields.Int(min=100, max=500, example=200) |
| 51 | + ) |
| 52 | +
|
| 53 | + # Averaged measurement action |
| 54 | + labthing.build_action( |
| 55 | + my_component.average_data, # Python function |
| 56 | + "/average", # URL to bind the action to |
| 57 | + description="Take an averaged measurement", |
| 58 | + args={ # How do we convert from the request input to function arguments? |
| 59 | + "n": fields.Int(description="Number of averages to take", example=5, default=5) |
| 60 | + }, |
| 61 | + ) |
| 62 | +
|
| 63 | +
|
| 64 | + # Start the app |
| 65 | + if __name__ == "__main__": |
| 66 | + from labthings.server.wsgi import Server |
| 67 | + Server(app).run() |
| 68 | +
|
| 69 | +
|
| 70 | +
|
| 71 | +For completeness of the examples, our ``PretendSpectrometer`` class code is: |
| 72 | + |
| 73 | +.. code-block:: python |
| 74 | +
|
| 75 | + import random |
| 76 | + import math |
| 77 | + import time |
| 78 | +
|
| 79 | + class PretendSpectrometer: |
| 80 | + def __init__(self): |
| 81 | + self.x_range = range(-100, 100) |
| 82 | + self.magic_denoise = 200 |
| 83 | +
|
| 84 | + def make_spectrum(self, x, mu=0.0, sigma=25.0): |
| 85 | + """ |
| 86 | + Generate a noisy gaussian function (to act as some pretend data) |
| 87 | + |
| 88 | + Our noise is inversely proportional to self.magic_denoise |
| 89 | + """ |
| 90 | + x = float(x - mu) / sigma |
| 91 | + return ( |
| 92 | + math.exp(-x * x / 2.0) / math.sqrt(2.0 * math.pi) / sigma |
| 93 | + + (1 / self.magic_denoise) * random.random() |
| 94 | + ) |
| 95 | +
|
| 96 | + @property |
| 97 | + def data(self): |
| 98 | + """Return a 1D data trace.""" |
| 99 | + return [self.make_spectrum(x) for x in self.x_range] |
| 100 | +
|
| 101 | + def average_data(self, n: int): |
| 102 | + """Average n-sets of data. Emulates a measurement that may take a while.""" |
| 103 | + summed_data = self.data |
| 104 | +
|
| 105 | + for _ in range(n): |
| 106 | + summed_data = [summed_data[i] + el for i, el in enumerate(self.data)] |
| 107 | + time.sleep(0.25) |
| 108 | +
|
| 109 | + summed_data = [i / n for i in summed_data] |
| 110 | +
|
| 111 | + return summed_data |
0 commit comments