MTOOLS

A collection of classes to:

• retrieve surface reflectance products (Maja S2, Maja Venus zip, HDF) as numpy array, accounting for cloud and edge masks;
• compute statistics from products, collection of products, or from comparison between two collections.

Status

version "1.0.3"

Testing coverage (with pytest):

----------- coverage: platform linux, python 3.7.7-final-0 -----------
Name                        Stmts   Miss  Cover   Missing
---------------------------------------------------------
common/Collection.py           65      5    92%   19, 60, 91-92, 106
common/Comparison.py           38      8    79%   15-19, 50-52
common/Product.py             172     21    88%   20-23, 54-60, 70-71, 92, 136-139, 242, 245, 258, 265
common/Roi.py                  78     12    85%   51-52, 57-58, 72, 74, 94-98, 112, 114, 122
common/test_Collection.py      21      0   100%
common/test_Comparison.py      17      0   100%
common/test_Product.py        170      0   100%
common/test_Roi.py            103      0   100%
common/test_utilities.py       84      0   100%
common/utilities.py            74     11    85%   65-66, 75, 89-91, 158, 172-175
---------------------------------------------------------
TOTAL                         822     57    93%

Python v-env:

• requirement.txt (pip)
• mtools.yml (conda)

Product

Product.Product

Generic class with the following methods:

• get_content_list(): retreive the content of a product
• find_band(string): get the filename from a pattern
• get_band(band, [scalef]): get a band as numpy array, optionally apply scale factor
• get_band_subset(band, roi=None, ulx=None, uly=None, lrx=None, lry=None, scalef=None): get a subset of band by passing either an Roi object of coordinates, optionally with scale factor

Product.Product_zip

Extends Product for zipped files.

Product.Product_zip_venus

Extends Product.Product_zip with venus specific methods and attributes.

Product.Product_dir_maja

Extends Product with Maja S2 specific methods and attributes.

Product.Product_hdf

Extends Product for HDF files.

Product.Product_hdf_acix

Extends Product.Product_hdf with ACIX reference S2 specific methods and attributes.

Roi

Roi.Roi_collection

A class to create a collection of ROI from pairs of coordinates described in a csv file and an extent.

Roi.Roi

An Roi class that can be passed to Product.get_band_subset()

Collection

Automatically finds any products in a given path and create collection instances for Comparison.

Comparison

Comparison.Comparison

Cornerstone class to compare two time series of products in collections. Provides facilities to find matching by date between two collections of products, and compute statistics for each band of each product of both collections. Create subclasses of Comparison for various scenarii.

Comparison.Comparison_acix

Extends Comparison.Comparison with specific methods for ACIX.

• one_by_one(): output RMSE for valid pixels between each band of each products of two collections passed to Comparison. A valid pixel is within image boundaries (from edge mask) and cloud-free (from cloud mask);
• flatten(): output one RMSE of the entire comparison.

ROISTATS

Lightweight utility to compute band statistics from zipped Venus products over user defined Regions of Interest.

Coordinates are expressed in UTM, ROI extent in meters.

Usage:

python roistats.py <full_path_produit.zip> <roi_definition.csv> [-v|e]

A ROI definition file must be in csv, with one line per ROI containing:

<int_roi_id>, <float_utmx>, <float_utmy>

With:

• int_roi_id: any user-defined int

• float_utmx: x (UTM) centre coordinate of the ROI

• float_utmy: y (UTM) centre coordinate of the ROI

Optional arguments:

-v, --verbose : set log file verbosity to DEBUG (defaults to INFO)

-e, --extent : set ROI extent (in meters, defaults to 100m). An ROI is a square of size extent*extent centered on (utmx, utmy)

For help:

roistats.py --help'

Requirements:

Dependency list provided as yaml conda environment file in:

mtools.yml