Starting example

Example 1: Simple annotation¶

This is a simple example to showcase the general usage of lmrtfy. It can be found in examples/starting_example/example1.py.

The two core concepts are the variable and result functions which annotate the inputs and outputs of the script. They are needed to create the profile which is used to create the API.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15

import numpy as np
from lmrtfy.annotation import variable, result # (1)!


x = variable(5, name="x", min=1, max=10) # (2)!
y = variable(np.linspace(0., 1., 101, dtype=np.float64), name="y", 
             min=-1., max=11., unit="m") # (3)!
z = variable("abc", name="z")

z1 = variable(["abc", "def"], name="z1")  # (4)!
z2 = variable(["abc", 1, 1.1], name="z2") 
z3 = variable({'a': "abc", 'b': 1}, name="z3") 

a = result(x * y, name="a") # (5)!
b = result(x * z, name="b")

1. The functions need to be imported from the lmrtfy library
2. The variable x has the local value 5 and can be between 1 and 10.
3. You can have numpy arrays as inputs
4. Lists and dictionaries work, too!
5. Results are similar to variables. They have a name and an expression that they will become.

Run python examples/starting/example1.py to create the profile needed for the deployment.

Deployment¶

To deploy the script run lmrtfy deploy examples/starting/example1.py --local

Call example1 from code¶

Now you can simply call catalog.example1() with the correct arguments, and you are good to go:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17

import time

from lmrtfy.functions import catalog

job = catalog.<your_namespace>example1(x=1,   # (1)!
                       y=[1, 2.0, 3.0],
                       z="foobar",
                       z1=["bar", "foo"],
                       z2=["foo", 1, 42],
                       z3={"foo": "bar", "bar": "foo"}
                       )

if job:
    while not job.ready:
        time.sleep(1)

    print(job.results)

1. <your_namespace> is your private namespace on LMRTFY, which is typically your nickname. Available namespaces are shown when importing catalog or when calling catalog.update().

if job: is currently required to ensure that you actually got a job object back from the function which would not be the case if the submission failed.

Calling example1 from the CLI¶

During the deployment you should have received a seven letter profile_id:

Profile_id to be used for requests: CgHUejl

We need the profile_id to submit a job from the CLI:

lmrtfy submit CgHUejL examples/starting/example1.json

If the JSON file has the correct inputs, in a valid range with correct units you will see that the job submission was successful. You will receive a ten-letter job ID.

INFO Job submission successful.
INFO Job-id: HgaUbcTFah

With this job_id you can now get the job results:

lmrtfy fetch HgaUbcTFah .  

That's all that is to it. Happy Hacking!