Skip to content

Latest commit

 

History

History
168 lines (124 loc) · 8.33 KB

Installation.md

File metadata and controls

168 lines (124 loc) · 8.33 KB
Raw

Table of Contents

  1. Necessary Software
  2. Step 1: Conda Set Up
  3. Step 2: Clone or Download Git Repo
  4. Step 3: Recommended Software
  5. Step 4: Optional Jupyter Notebook Set Up
  6. Example Datasets
  7. Optional Software
  8. Troubleshooting
  9. Conda Basics

Necessary Software

  • Miniconda is a free, light-weight version of conda. It will be used to install most essential command line based software. Details on how to install miniconda are in the section below

  • Workshop materials will be hosted on the github repo: https://github.com/mattbeckley/PDALIntro Updates to the workshop will be pushed to this site, so make sure to sync with the latest version before the workshop (see section: Clone or Download Git Repo)

  • PDAL. The workshop will follow the markdown tutorials using a command line version of PDAL. Users are expected to have a working version of PDAL so that they can copy and paste commands from the tutorial into their PDAL installation. It is HIGHLY recommended that users use conda to install PDAL. Details on how to do this are below

  • GDAL. Portions of the tutorial will utilize GDAL to grid point clouds, look at metadata, perform raster math, etc. It is recommended that users use conda to install GDAL. Details on how to do this are below.

Step 1: Conda Set Up

Conda is an open source package management system and environment management system that runs on Windows, macOS and Linux. Conda quickly installs, runs, and updates packages and their dependencies. Miniconda is a free, light-weight version of conda. It is recommended to use miniconda to install PDAL and other packages for this workshop.

Download the conda installer for your OS setup. https://docs.conda.io/en/latest/miniconda.html

  • Note if using Mac OSx it is probably easiest to use the .pkg file. Also, if using a newer mac with M1 chips make sure to select that version (i.e. Miniconda3 macOS Apple M1 64-bit pkg) and NOT the older intel-based option (Miniconda3 macOS Intel x86 64-bit pkg)

After installing conda, create an isolated workspace for this workshop. Windows users: Once installed use the "Anaconda Prompt (miniconda)" available under Programs > Miniconda3 to utilize conda in the terminal.

conda create --name pdalworkshop
conda activate pdalworkshop

Install PDAL, GDAL, jq via conda:

conda install -c conda-forge pdal jq gdal

To verify your installation of PDAL and GDAL, issue the following commands and you should see the version numbers printed to the terminal:

pdal --version
gdalinfo --version

Step 2: Clone or Download Git Repo

  • All the workshop materials: tutorials, sample data, sample pipelines, etc are in the repo: https://github.com/mattbeckley/PDALIntro

  • Users familiar with git should clone this repo

  • If you are unfamiliar with git or unable to install it, you can download a zip file directly from Github repo so that workshop materials are available locally:

    Github Download Zip

Step 3: Recommended Software

Step 4: Optional Jupyter Notebook Set Up

Note For this workshop, use of the notebooks is optional and is designed for use with bash shell (NOT for Windows).

A bash notebook with PDAL commands will be available. Users who wish to use this notebook will need to install the bash_kernel. To enable a bash kernel in a jupyter notebook environment:

conda activate pdalworkshop
conda install -c conda-forge notebook nb_conda_kernels `

FOR NON-WINDOWS USERS ONLY: pip install bash_kernel ; python -m bash_kernel.install

  • To run the jupyter notebook, run the following command from the notebook location:

jupyter notebook

Example Datasets

  • There are some example LAZ files under the data directory of the git repo. However, users are also encouraged to use their own data, or download data from OpenTopography. Note for the workshop, it is best to work with smaller files to keep processing times short. As a rule of thumb, keeping datasets below 5 million points is recommended for quick processing times.

Optional Software

  • As time permits, the workshop will highlight a simplified workflow of how to access USGS 3DEP data from AWS for those who are unfamiliar with working with Jupyter Notebooks.

  • Users who are comfortable with Jupyter Notebooks are encouraged to explore OpenTopography notebooks for working with USGS 3DEP data. There are a series of notebooks here: https://github.com/OpenTopography/OT_3DEP_Workflows/tree/main/notebooks. As an introduction, it is recommended to start with: 01_3DEP_Generate_DEM_User_AOI.ipynb which will explain how to download USGS 3DEP data, and create a DEM. Follow the instructions for "Option 2: Local Installation", specifically:

#create new working directory for 3DEP work...
mkdir 3DEP
cd 3DEP
git clone https://github.com/OpenTopography/OT_3DEP_Workflows
cd OT_3DEP_Workflows
conda env create -n 3dep --file environment.yml
conda activate 3dep

Troubleshooting

  • If having trouble with a conda environment, sometimes it is best to start with a fresh workspace:
#remove old workspace
conda env remove --name pdalworkspace

#Create a new, fresh workspace and activate it
conda create --name pdalworkshop2
conda activate pdalworkshop2
  • Occasionally there may be issues when installing a bunch of programs in a single command. If you encounter errors, it may be worthwhile splitting up the installation, particularly for large packages (e.g. GDAL, PDAL):
#If there are issues, sometimes it helps to run the installs as separate commands.
conda install -c conda-forge pdal  
conda install -c conda-forge gdal 
conda install -c conda-forge jq notebook nb_conda_kernels
  • Your path should be set to include conda automatically in your .profile or .bashrc. However, if there are problems starting conda, you may need to explicitly set your PATH system variable by putting this in your .profile or .bashrc, depending on your set up:
export PATH="/home/username/miniconda/bin:$PATH"

Conda Basics

  • Below are some basic, commonly used conda commands

  • To verify that conda is installed and running on your system:

conda --version
  • To get into a workspace:
conda activate pdalworkshop
  • Similarly, to exit:
conda deactivate
  • To remove a workspace:
conda env remove --name sample_workspace
  • To get a list of the environments created on your system:
conda info --envs
  • To clone an existing environment (in this example, "workspace"):
conda create --name workspace_clone --clone workspace
  • To search for a package to install:
conda search lftp
  • To search for a package on a specific conda channel:
conda search -c conda-forge lftp
  • To install a specific version of software:
conda create -n gdal244 -c conda-forge gdal==2.4.4