atomai 0.2.6

Deep and machine learning for atom-resolved data

AtomAI

What is AtomAI?

AtomAI is a simple Python package for machine learning-based analysis of experimental atom-resolved data from electron and scanning probe microscopes, which doesn't require any advanced knowledge of Python (or machine learning). It is the next iteration of the AICrystallographer project.

How to use it?

AtomAI has two main modules: atomnet and atomstat. The atomnet is for training neural networks (with just one line of code) and for applying trained models to finding atoms and defects in image data (which also takes a single line of code). The atomstat allows taking the atomnet predictions and performing the statistical analysis on the local image descriptors associated with the identified atoms and defects (e.g., principal component analysis of atomic distortions in a single image or computing gaussian mixture model components with the transition probabilities for movies).

Here is an example of how one can train a neural network for atom/defect finding with essentially one line of code:

from atomai import atomnet

# Here you load your training data
dataset = np.load('training_data.npz')
images_all = dataset['X_train']
labels_all = dataset['y_train']
images_test_all = dataset['X_test']
labels_test_all = dataset['y_test']

# Train a model
trained_model = atomnet.trainer(
    images_all, labels_all, 
    images_test_all, labels_test_all,
    training_cycles=500).run()   

Trained models can be used to find atoms/defects in the previously unseen (by a model) experimental data:

# Here you load new experimental data (as 2D or 3D numpy array)
expdata = np.load('expdata-test.npy')

# Get model's "raw" prediction, atomic coordinates and classes
nn_input, (nn_output, coordinates) = atomnet.predictor(expdata, trained_model, refine=False).run()

One can then perform statistical analysis using the information extracted by atomnet. For example, for a single image, one can identify domains with different ferroic distortions:

from atomai import atomstat

# Get local descriptors
imstack = atomstat.imlocal(nn_output, coordinates, crop_size=32, coord_class=1)

# Compute distortion "eigenvectors" with associated loading maps and plot results:
nmf_results = imstack.imblock_nmf(n_components=4, plot_results=True)

For movies, one can extract trajectories of individual defects and calculate the transition probabilities between different classes:

# Get local descriptors (such as subimages centered around impurities)
imstack = atomstat.imlocal(nn_output, coordinates, crop_size=32, coord_class=1)

# Calculate Gaussian mixture model (GMM) components
components, imgs, coords = imstack.gmm(n_components=10, plot_results=True)

# Calculate GMM components and transition probabilities for different trajectories
traj_all, trans_all, fram_all = imstack.transition_matrix(n_components=10, rmax=10)

# and more

Quickstart: AtomAI in the Cloud

The easiest way to start using AtomAI is via Google Colab

1. Use AtomAI to train a deep NN for atom finding

2. Analyze distortion domains in a single atomic image

3. Analyze trajectories of atomic defects in atomic movie - TBA

4. Prepare training data from experimental image with atomic coordinates (beta)

5. Atom finding using deep ensembles for uncertainty quantification (advanced)

Installation

First, install PyTorch. Then, install AtomAI via

pip install atomai