projpicker 0.1.4

ProjPicker (projection picker) allows the user to select all projections whose extent completely contains given points, polylines, polygons, and bounding boxes. The goal is to make it easy and visual to select a desired projection by location.

ProjPicker

ProjPicker (projection picker) allows the user to select all coordinate reference systems (CRSs) whose extent completely contains given points, polylines, polygons, and bounding boxes. The goal is to make it easy and visual to select a desired projection by location. This project was motivated by a GRASS GIS feature request. It is a work in progress; join discussions. See also the command-line and API documentation.

Branches

main

The single Python script projpicker.py provides the CLI and API for ProjPicker.

rtree-oop

The rtree-oop branch utilizes the rtree module in an object-oriented programming (OOP) way. This branch supports point intersection. The rtree spatial indexing may not be able to handle reversed west and east longitudinal boundaries that cross the antimeridian.

Requirements

Tested with proj.db from pyproj 3.1.0 in Python 3.7.2

Requires the PROJ database (e.g., /usr/share/proj/proj.db) only for recreating the provided projpicker.db, if you want, and the following standard modules:

• collections
• argparse
• os
• sys
• sqlite3
• re
• math
• json
• pprint

To recreate projpicker.db, the pyproj module is required.

Installation

• GitHub repository
• Python package

pip3 install projpicker

# or if you're not a root
pip3 install --user projpicker

# to install development versions
pip3 install --pre projpicker

# or if you're not a root
pip3 install --pre --user projpicker

Creating the ProjPicker database

This step is optional because projpicker.db generated from pyproj 3.1.0 is shipped with the module by default. Run this step only when you want to recreate this database from your version of PROJ.

From the shell,

projpicker -c

From Python,

import projpicker as ppik
ppik.create_projpicker_db()

Supported coordinate formats

The following geometry file points.txt contains 11 identical points:

################################
# decimal degrees and separators
################################
34.2348,-83.8677		# comma
34.2348		-83.8677	# whitespace

####################################################
# degree, minute, and second symbols
# degree: ° (U+00B0, °, alt+0 in xterm), o, d
# minute: ' (U+0027, '), ′ (U+2032, ′), m
# second: " (U+0022, "), ″ (U+2033, ″),
#	  '' (U+0027 U+0027, ' '), s
####################################################
34.2348°	-83.8677°	# without minutes, seconds, and [SNWE]
34°14.088'	-83°52.062'	# without seconds and [SNWE]
34°14'5.28"	-83°52'3.72"	# without [SNWE]
34.2348°N	83.8677°W	# without minutes and seconds
34°14.088'N	83°52.062'W	# without seconds
34°14'5.28"N	83°52'3.72"W	# full
34°14′5.28″N	83°52′3.72″W	# full using U+2032 and U+2033
34o14'5.28''N	83o52'3.72''W	# full using o' and ''
34d14m5.28sN	83d52m3.72sW	# full using dms

Running projpicker -p -i points.txt will generate:

[[34.2348, -83.8677],
 [34.2348, -83.8677],
 [34.2348, -83.8677],
 [34.2348, -83.8677],
 [34.2348, -83.8677],
 [34.2348, -83.8677],
 [34.2348, -83.8677],
 [34.2348, -83.8677],
 [34.2348, -83.8677],
 [34.2348, -83.8677],
 [34.2348, -83.8677]]

Querying points

Results are sorted by area to highlight local CRSs.

From the shell,

# read latitude and longitude separated by a comma or whitespaces from
# arguments
projpicker 34.2348,-83.8677 "33.7490  84.3880W"

# read latitude and longitude from stdin
projpicker <<EOT
# query points
34.2348		83°52'3.72"W	# UNG Gainesville Campus
33°44'56.4"	-84.3880	# Atlanta
EOT

From Python,

import projpicker as ppik
bbox = ppik.query_points([[34.2348, -83.8677], [33.7490, -84.3880]])
ppik.print_bbox(bbox)

Querying polylines or polygons

From the shell,

# read latitude,longitude from arguments
projpicker -g poly -- -10,0 10,0 10,10 10,0 , 10,20 30,40

# read latitude,longitude from stdin
projpicker -g poly <<EOT
# poly 1
# south-west corner
10S,0
10,0	# north-west corner
	# this comment-only line doesn't start a new poly
# north-east corner
10	10
# north-west corner
10	0
poly 2	# "poly 2" is neither a comment nor a point, so we start a new poly
10	20
30	40
EOT

From Python,

import projpicker as ppik
bbox = ppik.query_polys([[[-10, 0], [10, 0], [10, 10], [10, 0]],
                         [[10, 20], [30, 40]]])
ppik.print_bbox(bbox)

Querying bounding boxes

From the shell,

# read south,north,west,east from arguments
projpicker -g bbox 0,0,10,10 20,20,50,50

# read south,north,west,east from stdin
projpicker -g bbox <<EOT
# region 1
0	0	10	10

# region 2
20	20	50	50
EOT

From Python,

import projpicker as ppik
bbox = ppik.query_bboxes([[0, 0, 10, 10], [20, 20, 50, 50]])
ppik.print_bbox(bbox)

Filtering

From the shell,

# projected CRSs only
projpicker 34.2348,-83.8677 33.7490,-84.3880 | grep "^projected_crs"

# CRSs in meter only
projpicker 34.2348,-83.8677 33.7490,-84.3880 | grep ",meter,"

From Python,

import projpicker as ppik

# sorted by area to find the most local CRS first
bbox = ppik.query_points([[34.2348, -83.8677], [33.7490, -84.3880]])

# projected CRSs only
bbox_proj = list(filter(lambda x: x.proj_table=="projected_crs", bbox))
ppik.print_bbox(bbox_proj)

# CRSs in meter only
bbox_meter = list(filter(lambda x: x.unit=="meter", bbox))
ppik.print_bbox(bbox_meter)

Finding missing projection information

The ProjPicker can be used to guess the projection of data whose spatial reference information is missing for some reason (e.g., a Shapefile with no PRJ file). For example, I created a Shapefile (Atlanta_GA.shp) for Atlanta, Georgia, and deleted its PRJ file to simulate missing metadata. Only the filename gives the user a hint about its geographic location, but without the PRJ file, novice GIS users can have difficulty finding the right projection and repairing the Shapefile. When the file is opened in a GIS, it will be located far away from the true data location because projected coordinates are treated as latitudes and longitudes.

Nigeria is definitely not the right location, so the user can search for the latitude and longitude of Atlanta, GA (33.7490°N,84.3880°W) and check the extent of the Shapefile from the layer properties (1323252,1374239,396255,434290 in SNWE):

As can be seen in the screenshot, the spatial reference is unknown. Let's use the ProjPicker to guess its projection:

projpicker -n xy bbox 1323252,1374239,396255,434290 latlon point 33.7490°N,84.3880°W | head -1

This command line does not print the header line (-n) with column names to print only the first selection that is the most localized CRS to the query information. It then switches to the xy coordinate system and specifies the extent as bbox 1323252,1374239,396255,434290 followed by a point in latitude and longitude for Atlanta latlon point 33.7490°N,84.3880°W. The head -1 pipe prints the first CRS only. Remember, ProjPicker outputs are sorted by area from the smallest (most local) to largest (most global). The final output looks like:

projected_crs,NAD27 / Georgia West,EPSG,26767,EPSG,6602,EPSG,2190,30.62,35.01,-85.61,-82.99,227321.736222316,1825636.8909584181,45969.582735703174,870089.0814069586,US foot,119521.02819197961

From Python,

import projpicker as ppik

bbox = ppik.query_mixed_geoms("xy bbox 1323252,1374239,396255,434290 latlon point 33.7490°N,84.3880°W")
# equivalent to
# bbox = ppik.query_mixed_geoms(['xy', 'bbox', [1323252.0, 1374239.0, 396255.0, 434290.0], 'latlon', 'point', [33.749, -84.388]])

ppik.print_bbox(bbox[0])

EPSG:26767 is actually the correct CRS:

TODO

1. GUI
   • ArcGIS Pro Toolbox for ArcGIS users including IESA students? Will be easier to implement because ArcGIS Pro provides nice pencil tools and mapping functionalities. We'll be able to see almost immediate returns.
   • Web (client-only)
   • Desktop
2. CRS hints
   • Crowdsourcing agency and product information?

Versioning

N(.N)*[{a|b|rc}N][.postN][.devN]

• PEP 440
• {a|b|rc|.dev}N towards and .postN away from the release
• Not fully compatible with semantic versioning
• Not using build numbers marching away from or towards a release, but check this interesting comment.

Sponsor

This project is kindly funded by the Institute for Environmental and Spatial Analysis (IESA) at the University of North Georgia (UNG).

License

Copyright (C) 2021 Huidae Cho and Owen Smith

This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program. If not, see <https://www.gnu.org/licenses/>.