Getting more precision with GeoPandas?

Question

Geopandas

Reading a shapefile with GeoPandas and printing its total bounds:

dataframe = gpd.read_file('example.shp')
print(dataframe.total_bounds)

Prints: [ 663590.5817 1541419.8307 724630.0589 1595869.4839]

PyQGIS

Doing the same with PyQGIS:

QgsApplication.setPrefixPath("/usr", True)
qgs = QgsApplication([], False)
qgs.initQgis()
layer = QgsVectorLayer('example.shp', 'layer1', 'ogr')
print(layer.extent())

Prints: <QgsRectangle: 663590.58169999998062849 1541419.83070000074803829, 724630.05890000029467046 1595869.48389999940991402>

---

As you can see doing it with PyQGIS gives you more precision than doing it with GeoPandas.

Is there any way to get the same precision with GeoPandas?

Answer 1

Use geopandas.options.display_precision to specify how many decimal places to be visualized when printing the dataframe.

Note that internally the precision is not rounded, this is just a matter if displaying the coordinates.

Example, default behaviour:

>>> import geopandas
>>> world = geopandas.read_file(geopandas.datasets.get_path('naturalearth_lowres'))
>>> world['centroid_column'].head()
0    POINT (163.85316 -17.31631)
1      POINT (34.75299 -6.25773)
2     POINT (-12.13783 24.29117)
3     POINT (-98.14238 61.46908)
4    POINT (-112.59944 45.70563)
Name: centroid_column, dtype: geometry

With 9 decimal places now:

>>> geopandas.options.display_precision = 9
>>>
>>> world['centroid_column'].head()
0    POINT (163.853164645 -17.316309426)
1      POINT (34.752989855 -6.257732429)
2     POINT (-12.137831112 24.291172960)
3     POINT (-98.142381372 61.469076145)
4    POINT (-112.599435912 45.705628002)
Name: centroid_column, dtype: geometry

---

Taken from geopandas official documentation.

Answer 2

Try using pandas (pd) settings: pd.set_option('max_colwidth', value)

look here for more info: https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.set_option.html

Answer 3

You can change the data type in your GDF schema. The fiona documentation describes that you can specify a data type in the schema with its length for most data types, and you can also specify the precision of a float dtype. If you change the types in the schema, the accuracy and length will be preserved when importing to a file of the desired format, for example, ESRI ShapeFile.There are examples in the fiona documentation scheme section, the type is indicated (for example, float, followed by : and length, and then put . and the accuracy is indicated. Example: 'float:32.5'. For a given number, the length of the integer part will be 31 characters (a comma is considered a symbol, add 1 to get 32 digits before the comma), and its precision after the comma will be five.

An answer with an example of Python code: https://gis.stackexchange.com/a/483104/243816