nishadhka

Machine learning libraries of Python in Docker

The docker tar.gz'd is 2.4 GB in size, download from here

background and installed libraries in the docker

to set up the new ubuntu docker image, it is 124MB in size,
```
sudo docker search ubuntu
sudo docker pull ubuntu
```

Docker and Jupyter notebooks, portal to all Python environments

Background

Docker provides operating system level virtualisation. It eases the setting up, cloning and distribution of fully functional GUI less Linux operating system. In python programming, the setting of library and distribution can be difficult in many cases if the libraries are inherently depended on OS and low level languages. This become bottle neck in learning and experimenting with Python. Docker removes this barrier by file like virtual environments of fully functional OS.

The current workshop gives an hands on training on setting up docker for major operating systems. We will be using nested virtualisation in windows OS and native application for Linux OS’s. The docker setup and consuming various data science oriented docker containers will be looked upon. After setting up the docker, we will be discussing on Jupyter notebooks, a perfect GUI for docker headless server situation.

Objectives

Introduce participants about the

	#!/bin/bash

	# usage: nbgrep 'pattern'

	SEARCHPATH='/home/workingnotesBack'

	# 'jq' technique lifted with gratitude
	# from https://gist.github.com/mlgill/5c55253a3bc84a96addf

	# Break on newlines instead of any whitespace

	from pykml import parser
	import pandas as pd

	filename='ref1.kml'
	with open(filename) as f:
	folder = parser.parse(f).getroot().Document.Folder

	plnm=[]
	cordi=[]
	for pm in folder.Placemark:

	import pandas as pd
	from shapely.geometry import Point, shape
	from geopandas import GeoDataFrame
	import fiona
	from collections import OrderedDict

	# a test to create shape file
	df=pd.read_csv('VNP14IMGTDL_NRT_South_Asia_24h.csv')
	geometry = [Point(xy) for xy in zip(df.longitude, df.latitude)]
	df = GeoDataFrame(df, geometry=geometry)

	from shapely import geometry

	ll=(84.945856,23.475927)
	ue=(86.505248, 23.871591)

	p1 = geometry.Point(ll[0],ll[1])
	p2 = geometry.Point(ue[0],ll[1])
	p3 = geometry.Point(ue[0],ue[1])
	p4 = geometry.Point(ll[0],ue[1])

	import rasterio.merge

	bounds=None
	res=None
	nodata=None
	precision=7

	def merge(input1,bounds, res, nodata, precision):
	import warnings
	warnings.warn("Deprecated; Use rasterio.merge instead", DeprecationWarning)