stacterm

This library is for displaying information (tables, calendars, plots, histograms) about STAC Items in the terminal. It takes as input a STAC ItemCollection (a GeoJSON FeatureCollection of STAC Items), either by specifying a filename or by piping output from another program.

Installation

Install from PyPi:

$ pip install stacterm

PySTAC and Pandas are required, along with two dependencies for rendering tables (termtables) and plots (plotext) in the terminal.

Usage

stacterm main usage is as a CLI progam stacterm. Use help to see options available:

$ stacterm -h
usage: stacterm [-h] {table,cal,hist,plot} ...

Terminal STAC

positional arguments:
  {table,cal,hist,plot}
    table               Output a table
    cal                 Output a calendar
    hist                Output a histogram
    plot                Output a plot

optional arguments:
  -h, --help            show this help message and exit

The detailed usage examples below are shown using a saved file, however stacterm can also read in stdin allowing other programs to pipe output to it, such as pystac-client.

$ export STAC_API_URL=https://earth-search.aws.element84.com/v0
$ stac-client search --intersects aoi.json --datetime 2020-07-01/2020-12-31 -c sentinel-s2-l2a-cogs landsat-8-l1-c1 | stacterm cal --label platform

All of the sub-commands in stacterm can take optional field names. A field name is:

• id: The ID of the Item
• date: The date portion of the Item's datetime field
• collection: The collection of the Item
• Any property

Tables

Use stacterm to display tabularized data from a saved ItemCollection.

$ stacterm table items.json

| id                                       | date       |
|------------------------------------------|------------|
| LC08_L1TP_026079_20201014_20201104_01_T1 | 2020-10-14 |
| LC08_L1TP_026079_20201115_20201210_01_T1 | 2020-11-15 |
| S2A_12JXQ_20201008_0_L1C                 | 2020-10-08 |

By default this is a markdown table (note the terminal will not render Markdown)

The fields displayed can be changed via the --fields keyword, and sorted via the --sort keyword.

$ stacterm table items.json --fields date eo:cloud_cover collection --sort eo:cloud_cover

| date       | eo:cloud_cover | collection           |
|------------|----------------|----------------------|
| 2020-10-13 | 0.0            | sentinel-s2-l1c      |
| 2020-10-13 | 0.0            | sentinel-s2-l2a      |
| 2020-10-13 | 0.0            | sentinel-s2-l2a-cogs |
| 2020-10-13 | 0.0            | sentinel-s2-l1c      |

The style of the table can also be changed via the --style keyword, although it will no longer be usable in a Markdown renderer. See termtables styles for list of styles.

$ stacterm table items.json --fields id date platform sentinel:grid_square --sort date --style thick

┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┳━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━┓
┃ id                                       ┃ date       ┃ platform    ┃ sentinel:grid_square ┃
┣━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╋━━━━━━━━━━━━╋━━━━━━━━━━━━━╋━━━━━━━━━━━━━━━━━━━━━━┫
┃ S2B_12JXR_20201003_0_L2A                 ┃ 2020-10-03 ┃ sentinel-2b ┃ XR                   ┃
┣━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╋━━━━━━━━━━━━╋━━━━━━━━━━━━━╋━━━━━━━━━━━━━━━━━━━━━━┫
┃ S2B_12JXQ_20201003_0_L1C                 ┃ 2020-10-03 ┃ sentinel-2b ┃ XQ                   ┃
┣━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╋━━━━━━━━━━━━╋━━━━━━━━━━━━━╋━━━━━━━━━━━━━━━━━━━━━━┫
┃ S2B_12JXQ_20201003_0_L2A                 ┃ 2020-10-03 ┃ sentinel-2b ┃ XQ                   ┃
┗━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┻━━━━━━━━━━━━┻━━━━━━━━━━━━━┻━━━━━━━━━━━━━━━━━━━━━━┛

Calendars

A UNIX-like calendar (see cal) is available to show dates of individual items. By default cal will use the field datetime (the collection datetime) and group Items by their Collection. These can be overridden by the --date_field and --label_field keywords. Note that the specified --date_field needs to be a date field, such as created or updated.

$ stacterm cal items.json --date_field created --label_field gsd

Histograms

Histograms can be created for any numeric field.

$ stacterm hist items.json eo:cloud_cover

Plots

Plots can be created with 1 or 2 numeric fields. If a single field it will be plotted against the scene number. The --sort keyword can control how to sort the data if plotting a single field.

$ stacterm plot items.json eo:cloud_cover --sort eo:cloud_cover

Development

There are a lot more options in the plotext library that could be surfaced here. Additionally, if support for datetimes in histograms and plots is added, stacterm could create temporal histograms, or plot quantities vs date.