dataframe-service 0.1.5

Pandas DataFrame Service (DFS).

Pandas DataFrame Service (DFS)

Features

* Simple, ~100 lines multi-threaded file cache implementation
* Key-value store for Panda DataFrames with basic index querying
* Fixed budget memory consumption w/ LRU eviction
* Supports updates on files and dataframes
* Simple TCP client/server interface w/ client-side connection pooling

Limitations

1. Currently does not support replication, though the file system can be (e.g. NAS)

Usage

Install

$ pip3 install dataframe-service 

Start server

$ dfs_server
Serving on 0.0.0.0 port 8000 with max memory 1073741824 at root directory <current dir>

Run DFS REPL

$ rlwrap dfs_repl
Welcome to DFS REPL
author: Brian Guarraci
repo  : https://github.com/briangu/dfs
crtl-c to quit

?> stats
{
  "memory": {
    "used": "0",
    "free": "1073741824",
    "max": "1073741824"
  }
}