20cmdingding/easy_memmap

Python memmap wrapper to share arrays easily between processes

Numpy Memmap easy wrapper for array manipulation

Python2/3 memmap wrapper to share arrays easily between processes in real time.

Installation

You can install easy_memmap from pip using

pip install easy_memmap

or using

pip install git+https://github.com/charlielito/easy_memmap

Numpy [d]types supported

The library accepts any type of the following supported by numpy

• np.bool
• np.int8, np.int16, np.int32, np.int64
• np.uint8, np.uint16, np.uint32, np.uint64
• np.float16, np.float32, np.float64

Usage

The program that shares the array can be

from easy_memmap import EasyMemmap
import numpy as np

m = EasyMemmap(mode="w", name="mytest")

data = np.zeros((480,640,4))
m.write(data)

And the programm that reads the array

from easy_memmap import EasyMemmap

m = EasyMemmap(mode="r", name="mytest")

data = m.read()

The name of the EasyMemmap object needs to be the same, otherwise it won't be able to find the data and it will return None

To see other usages streaming in real time images from a web_camera, see examples