Access to Fiona's crs_wkt and/or meta attributes when using read_file()

I have a shapefile which I open in geopandas using read_file(). It is in GDA94 mga54 (epsg: 28354). In geopandas it will print the crs attribute as

import geopandas as gpd
source_gdf = gpd.read_file('c:\data\test\boxes_gda94mga54.shp')
print source_gdf.crs
{u'zone': 54, u'ellps': u'GRS80', u'no_defs': True, u'proj': u'utm', u'units': u'm', 'wktext': True, u'south': True}`

When opening in Fiona and printing the coordinate information using the meta attribute it produces

import fiona,pprint
source_fio = fiona.open('c:\data\test\boxes_gda94mga54.shp')
pprint.pprint (source_fio.meta)
{'crs': {u'ellps': u'GRS80',
         u'no_defs': True,
         u'proj': u'utm',
         u'south': True,
         u'units': u'm',
         u'zone': 54},
 'crs_wkt': u'PROJCS["GDA94_MGA_zone_54",GEOGCS["GCS_GDA_1994",DATUM["Geocentric_Datum_of_Australia_1994",SPHEROID["GRS_1980",6378137,298.257222101]],PRIMEM["Greenwich",0],UNIT["Degree",0.017453292519943295]],PROJECTION["Transverse_Mercator"],PARAMETER["latitude_of_origin",0],PARAMETER["central_meridian",141],PARAMETER["scale_factor",0.9996],PARAMETER["false_easting",500000],PARAMETER["false_northing",10000000],UNIT["Meter",1]]',
 'driver': u'ESRI Shapefile',
 'schema': {'geometry': 'Polygon',
            'properties': OrderedDict([(u'Id', 'int:10'), (u'float_area', 'float:13.11'), (u'double_Len', 'float:19.11'), (u'short_id', 'int:10'), (u'Long_vert', 'int:10'), (u'Date_Creat', 'date'), (u'part_type', 'str:50')])}}

If I then save the file using geopandas to_file() without transforming/re-projecting then look at projection attributes of the saved file using Fiona’s meta again it produces a different crs_wkt result when compared with the above.

source_gdf.to_file('c:\data\test\boxes_gda94mga54_copy.shp',driver="ESRI Shapefile")

new_fio = fiona.open('c:\data\test\boxes_gda94mga54_copy.shp')
pprint.pprint (new_fio.meta)
'crs': {u'ellps': u'GRS80',
         u'no_defs': True,
         u'proj': u'utm',
         u'south': True,
         u'units': u'm',
         u'zone': 54},
 'crs_wkt': u'PROJCS["UTM_Zone_54_Southern_Hemisphere",GEOGCS["GCS_GRS 1980(IUGG, 1980)",DATUM["unknown",SPHEROID["GRS80",6378137,298.257222101]],PRIMEM["Greenwich",0],UNIT["Degree",0.017453292519943295]],PROJECTION["Transverse_Mercator"],PARAMETER["latitude_of_origin",0],PARAMETER["central_meridian",141],PARAMETER["scale_factor",0.9996],PARAMETER["false_easting",500000],PARAMETER["false_northing",10000000],UNIT["Meter",1]]',
 'driver': u'ESRI Shapefile',
 'schema': {'geometry': 'Polygon',
            'properties': OrderedDict([(u'Id', 'int:9'), (u'float_area', 'float:24.15'), (u'double_Len', 'float:24.15'), (u'short_id', 'int:9'), (u'Long_vert', 'int:9'), (u'Date_Creat', 'str:80'), (u'part_type', 'str:80')])}}

Is there a way when using geopandas read_file() to store or access the crs_wkt or meta attribute that could then then be used as a parameter in to_file() to correctly assign the original projection to the output file?

[martinfleis]

This seems to be fixed by using pyproj.CRS under the hood to store CRS information. Roundtrip correctly returns GDA94 / MGA zone 54, not UTM_Zone_54_Southern_Hemisphere anymore.

[cratcliff]

Thanks @jdmcbr and @jorisvandenbossche for the advice. I have tested the workaround and it does work.

I still think it would be beneficial to have access to a crs_wkt (and maybe epsg as well) as a reference and continue with the proj4 string for conversions. Its easy enough to convert epsg & crs_wkt to proj4 by using GDALs SpatialReference methods which is a dependency of Fiona. This will then provide a more detailed description of the data frames coordinate system to the user when it changes or becomes disconnected from the source.

I am keen to use pandas and geopandas for a project i’m working on which will manipulate and analyze data coming from various sources (shapefiles and csv’s). The analysis will include several datasets and projections to perform multiple geometry operations (buffering, overlays etc) and column calculations, and I will end up with multiple geodataframes so it would be good if the crs_wkt attribute were attached to the data frame rather than managed through separate variables or a custom class.