TSTools is a plugin for QGIS (version 2.4+) that helps visualize remote sensing time series by linking time series dataset models (objects that describe and characterize the time series) with user interface tools designed to harmonize the spatial and temporal dimensions of these large datasets.
Read the Quickstart to see the plugin in action
If you feel like TSTools has made a contribution to your research, please consider citing it using the plugin using the Digital Object Identifier (DOI) tracked by Zenodo:
An example dataset for this plugin is located here: https://github.com/ceholden/landsat_stack
This plugin has not been uploaded to the main QGIS plugin repository so installation will need to be done manually. Please make sure you have the required dependencies
One of the easiest ways to install TSTools, if you have the required dependencies, is to manually copy a "compiled" copy of the plugin into your QGIS plugins folder.
In most cases, the QGIS Python plugins folder will be located in your home directory within the ".qgis2/python/plugins" folder. Any plugins you have installed previously will be located here. For more information, see this excellent answer on Stack Exchange.
- Install QGIS and the required Python libraries (see requirements section below)
- Download the file "tstools.zip" from a release of TSTools on Github.
- Unzip the ZIP file to find the "tstools" folder.
- Copy this "tstools" folder into your QGIS Python plugins directory (see above for where this is located)
To enable the plugin, continue following the instructions for enabling the plugin.
For anyone interested in developing the plugin or for those working in *nix environments, the easiest way to install the plugin is to clone this repository via git and to compile and deploy the plugin:
git clone https://github.com/ceholden/TSTools.git
cd TSTools/tstools/
make derase
make clean
make
make deploy
To enable the plugin, continue following the instructions for enabling the plugin.
A quick way of installing and using TSTools is to utilize the included setup script for the Vagrant technology. Vagrant enables users to quickly and reproducibly configure and create lightweight virtual machines. I have included a Vagrantfile inside vagrant/ that sets up a Ubuntu "Xenial Xerus" 16.04 Linux virtual machine with TSTools and all pre-requisites installed.
To run TSTools using Vagrant, install Vagrant for your platform from their downloads page:
http://www.vagrantup.com/downloads
Installation instructions are available here.
You will also need software to run the virtual machine, or a "provider" as Vagrant calls it. I recommend VirtualBox because it works well, is cross-platform, and is free and open-source. More providers and instructions for using these providers is available on Vagrant's documentation page.
Once you have Vagrant and a provider installed, you can run the Vagrant machine as follows:
# Navigate into the folder and vagrant up!
cd vagrant/
vagrant upOnce the virtual machine has been downloaded, configured, and provisioned, you can connect to it and launch QGIS via SSH:
vagrant ssh
qgisThat's it! You can suspend, halt, or destroy (delete) the virtual machine when you're done using these as the <command> in vagrant <command>.
For more information about Vagrant and how to use the technology, check out their "Getting Started" section within their documentation page.
You may also wish to avoid any setup and begin using TSTools using a pre-built image with everything already installed. The disadvantage to this solution is that VM images are likely to be out of date because it is difficult to update the VM image as frequently as one would like.
You may download an image of Ubuntu 16.04 "Xenial Xerus" from my university's
FTP server at the URL below. It has been exported as a
VirtualBox appliance
and can import the image according to Oracle's instructions.
Instructions are also available in the readme file on the FTP server:
http://ftp-earth.bu.edu/public/ceholden/TSTools/
Once TSTools is installed, follow these steps to enable it:
- Launch QGIS and open the Plugin Manage dialog (Plugins menu -> Manage and Install Plugins)
- Check the box next to "TSTools" to enable the plugin
Two new icons will be added to the plugins toolbar. These icons have the letters "TS" in capital red colored letters. To initialize a timeseries dataset within the plugin, click the icon without the crosshair symbol. Point this dialog to your timeseries and configure any additional options before clicking "Okay". To retrieve the timeseries for any given pixel, add an image from your timeseries to QGIS using the "Images" tab and click the "TS" icon with the crosshairs to replace your current map tool with the "TSTools" map tool.
As this is a QGIS plugin, QGIS, GDAL, and Python 2.7+ are, of course, required. Most, if not all, installations of QGIS also provide a copy of numpy and matplotlib so most users should be able to use TSTools "out of the box".
matplotlib>=1.4.0
numpy>=1.8.0
GDAL>=1.10.0
The following are additional, optional dependencies:
palettable>=2.1.1
scandir
The palettable package provides better colormap support for the plots. scandir speeds up the process of the plugin finding timeseries data spread across many files and directories. markdown2 enables stylistic parsing of timeseries driver information formatted in Markdown.
CCDC Results Reader- For reading Continuous Change Detection and Classification (CCDC) results:
scipy>=0.12.0
- For reading Continuous Change Detection and Classification (CCDC) results:
YATSM CCDCesuqe Timeseries, and otherYATSM *drivers- For live plotting with YATSM (including CCDC-esque clone), you need to install the
YATSMpackage. See instructions at the following repository location:
- For live plotting with YATSM (including CCDC-esque clone), you need to install the
To help develop this plugin, you will need QGIS, Python, and the Qt developer tools for Python (for building). The Qt dependencies are available on Ubuntu in the pyqt4-dev-tools package.