anndata_quickstart_sparse_matrix_init_slicing_metadata.py

1import numpy as np
2import pandas as pd
3import anndata as ad
4from scipy.sparse import csr_matrix
5
6# 1. Initialize data
7counts = csr_matrix(np.random.poisson(1, size=(100, 2000)), dtype=np.float32)
8adata = ad.AnnData(counts)
9
10# 2. Add observation and variable metadata
11adata.obs_names = [f"Cell_{i:03d}" for i in range(adata.n_obs)]
12adata.var_names = [f"Gene_{i:03d}" for i in range(adata.n_vars)]
13
14# 3. Add unstructured annotation
15adata.obs["cell_type"] = np.random.choice(["B", "T", "Monocyte"], size=adata.n_obs)
16
17# 4. Slicing and Accessing
18# Get the first 5 cells and specific genes
19subset = adata[:5, ["Gene_000", "Gene_001"]]
20
21print(adata)
22print(subset)