Lakota

Lakota is a columnar storage solution for timeseries.

Lakota organises reads and writes through a changelog inspired by Git. This changelog provides: historisation, concurrency control and ease of synchronisation across different storage backends.

Quickstart

Install with pip install lakota

You can then run:

from lakota import Repo, Schema

ts_schema = Schema(timestamp="timestamp*", value="float")
repo = Repo("my-data-folder")  # or Repo("s3://my-s3-bucket")
clct = repo.create_collection(ts_schema, "temperature")
series = clct.series('Brussels')
df = {
    "timestamp": [
        "2020-01-01T00:00",
        "2020-01-02T00:00",
        "2020-01-03T00:00",
        "2020-01-04T00:00",
    ],
    "value": [1, 2, 3, 4],
}

series.write(df)
df = series[:'2020-01-03'].df()
print(df)
# ->
#    timestamp  value
# 0 2020-01-01    1.0
# 1 2020-01-02    2.0
# 2 2020-01-03    3.0

See lakota.repo and lakota.schema on how to create a repository and define collections.

See lakota.collection and lakota.series on how to create series and read/write data.

lakota.commit and lakota.changelog document how data is organised into files and directories.

"""
# Lakota

Lakota is a columnar storage solution for timeseries.

Lakota organises reads and writes through a changelog inspired by
Git. This changelog provides: historisation, concurrency control and
ease of synchronisation across different storage backends.


## Quickstart

Install with `pip install lakota`

You can then run:

``` python
from lakota import Repo, Schema

ts_schema = Schema(timestamp="timestamp*", value="float")
repo = Repo("my-data-folder")  # or Repo("s3://my-s3-bucket")
clct = repo.create_collection(ts_schema, "temperature")
series = clct.series('Brussels')
df = {
    "timestamp": [
        "2020-01-01T00:00",
        "2020-01-02T00:00",
        "2020-01-03T00:00",
        "2020-01-04T00:00",
    ],
    "value": [1, 2, 3, 4],
}

series.write(df)
df = series[:'2020-01-03'].df()
print(df)
# ->
#    timestamp  value
# 0 2020-01-01    1.0
# 1 2020-01-02    2.0
# 2 2020-01-03    3.0
```

See `lakota.repo` and `lakota.schema` on how to create a repository
and define collections.

See `lakota.collection` and `lakota.series` on how to create series
and read/write data.

`lakota.commit` and `lakota.changelog` document how data is organised
into files and directories.
"""

from .changelog import *
from .collection import *
from .frame import *
from .pod import *
from .repo import *
from .schema import *
from .series import *

__version__ = "0.7.1"

Sub-modules

lakota.batch

lakota.changelog

lakota.cli

Example usage …

lakota.collection

Read and Writes Series …

lakota.commit

The Commit class is responsible to structure the content of a commit file. A commit is like a sorted dataframe with the following columns: label, …

lakota.frame

lakota.http_pod

lakota.pod

The POD class implement low-level access to different storage. The from_uri method allow to instanciate a POD object based on a uri. The supported …

lakota.repo

The Repo class manage the organisation of a storage location. It provides creation and deletion of collections, synchronization with remote …

lakota.s3_pod

lakota.schema

lakota.series

lakota.server

The server sub-module implement a flask application that expose The main methods of POD. It can be launched from the cli like this: …

lakota.sexpr

The AST (abstract syntax tree) implement parsing and evaluation of s-expressions …

lakota.utils