Releases: MAAP-Project/gedi-subsetter
Releases · MAAP-Project/gedi-subsetter
0.9.0
0.8.0
Fixed
- Remove hard-coded MAAP API host value in the scripts
bin/algo/describe
,bin/algo/delete
, andbin/algo/register
lettingmaap-py
make use ofMAAP_API_HOST
environment variable (#85)
Changed
- Obtain AWS S3 credentials via a role using the EC2 instance metadata rather than via the
maap-py
library
(#14) - Log messages with timestamps in ISO 8601 UTC combined date and time representations with milliseconds
(#72) - Read granule files directly from AWS S3 instead of downloading them (#54)
- Optimize AWS S3 read performance to provide ~10% speed improvement (on average) over downloading files by tuning the
default_cache_type
,default_block_size
, anddefault_fill_cache
keyword arguments to thefsspec.url_to_fs
function
(#77) - Set default granule
limit
to 100000. Although this is not unlimited, it effectively behaves as such because all of the supported GEDI collections have fewer granules than this limit. (#69) - Set default job queue to
maap-dps-worker-32vcpu-64gb
to improve performance by running on 32 CPUs (#78) - Succeed even when the result is an empty subset (#79)
- Upgrade to Python 3.12
Added
0.7.0
Added
- #57 Users may choose to profile their jobs by specifying command-line options for the
scalene
profiling tool. Seedocs/MAAP_USAGE.md
for more information. - #44 Granule download failures are now retried up to 10 times to reduce the likelihood that subsetting will fail due to a download failure.
- #56 The
bin/subset
script now captures output tostderr
and writes it to the log file namedgedi-subset.log
. When a job succeeds, the log file will appear in the job's output directory. Otherwise, it will appear in the jobs triage directory. - #65 All supported GEDI collections are now cloud-hosted, and granules are now downloaded from the cloud rather than from DAAC servers.
0.6.2
Fixed
- Updated to use v3.1.3 of maap-py in environment-maappy.yml. Previous versions of maap-py were referencing the deprecated MAAP Query Service API endpoint.
Full Changelog: 0.6.1...0.6.2
0.6.1
0.6.0
Fixed
- #40 All geometries in an AOI (area of interest) file are now used for granule selection and subsetting. Previously, only the first geometry was used, resulting in a much smaller subset than expected, in cases where the AOI contains multiple geometries.
- #41 Granules without a download link in their metadata are now skipped. Previously, encountering such granules would cause a job failure, due to being unable to download a file without having a download link.
- #42 Granules with metadata containing multiple boundaries within the horizontal spatial domain are now supported. In such cases, a single boundary is obtained by taking the union of the individual boundaries. If the result intersects with the AOI, then the granule is included in the subset. Previously, although rare, such granule metadata would cause a job failure.
Changed
- Upgraded Python to version 3.11 to take advantage of the addition of fine-grained error locations in tracebacks to help with debugging errors.
- The
beam
column is no longer automatically included in the output file. If you wish to include thebeam
column, you must specify it explicitly in thecolumns
input. - The default value for
limit
was reduced from 10000 to 1000. The AOI for most subsetting operations are likely to incur a request for well under 1000 granules for downloading, so a larger default value might only lead to longer CMR query times.
Added
- #38: Temporal filtering is now supported, such that specifying a temporal range will limit the granules downloaded from the CMR, pulling only granules obtained within the specified range. See
README.md
for more information. - Added an input parameter named
output
to allow user to specify the name of the output file, rather than hard-code the name togedi-subset.gpkg
. SeeREADME.md
for more information.
0.5.0
0.4.0
0.3.0
Fixed
- #5: Nested variables must now be specified by path relative to each BEAM group. This not only avoids ambiguity for variables of the same name (but different paths), but also makes a variable's location explicit.
- #17: Granule files that cannot be successfully read are skipped, rather than causing job failure. Offending files are retained to facilitate analysis.
Added
- #1 User must specify values for
lat
andlon
as inputs, allowing the user to choose which lat/lon datasets are used. - #2: User must specify
doi
as an input, now allowing subsetting of L2A as well as L4A data. - #7: Columns from 2D variables can be selected.
- #8: Specifying a query is now optional, to allow selecting all rows for specified columns.
0.2.7
Added
- Promoted the GEDI Subsetting algorithm to this repository from the MAAP-Project/maap-documentation-examples repository. This
0.2.7
version replicates thegedi-subset-0.2.7
version released from that repository.