A client library for making LaTiS requests in Python.

We do not yet publish to PyPI so you need to install from Git:

python3 -m pip install git+https://github.com/lasp/latis-python-client.git

Use the read_data function to make a request to LaTiS and get the results as a Pandas DataFrame.

The read_data function expects the following arguments:

• base_url: The base URL of a LaTiS instance. The base URL includes the host and the path through /dap (for LaTiS 2) or /dap2 (for LaTiS 3).
• dataset: The identifier of a dataset.
• start_time: The start time of the request, in ISO-8601 format.
• end_time: The end time (exclusive) of the request, in ISO-8601 format.

from latis.client import read_data

df = read_data(
    "https://lasp.colorado.edu/lisird/latis/dap2",
    "bremen_composite_mgii",
    "2026-01-01",
    "2026-01-02"
)

You can specify an API key for instances that require one.

df = read_data(
    "https://lasp.colorado.edu/lisird/latis/dap2",
    "bremen_composite_mgii",
    "2026-01-01",
    "2026-01-02",
    api_key="<api key>"
)

You can specify a username and password for instances that require basic auth.

df = read_data(
     "https://lasp.colorado.edu/lisird/latis/dap2",
     "bremen_composite_mgii",
     "2026-01-01",
     "2026-01-02",
     username="<username>",
     password="<password>"
 )

For queries more complicated than time selections, you can optionally specify an additional DAP2 query fragment that will be appended to the query sent to LaTiS. It must be URL-encoded.

df = read_data(
    "https://lasp.colorado.edu/lisird/latis/dap",
    "nnl_hires_ssi_P1D",
    "2025-01-01",
    "2025-01-02",
    query="wavelength%3E=400&wavelength%3C500&convert(time,days%20since%201858-11-17)&replace_missing(-99)"
)

Multiple datasets can be joined by passing a list or tuple of dataset identifiers to read_data. This is only supported by LaTiS 3 instances (those with URLs ending in /dap2).

df = read_data(
  "https://example.org/latis/dap2",
  ["dataset1", "dataset2"],
  "yyyy-mm-dd",
  "yyyy-mm-dd"
)