AttributeError: 'pandas' object has no attribute 'gen'-CodePudding

I am pretty new in programming. I am trying to get graphs from polygons from a shp-file. This code runs with another file. But in this file, some of the graph-names don't work. An error occurs that the 'pandas' object has no attribute 'gen'.

    import geopandas as gpd
    import osmnx as ox
    path_to_folder = r'/Users/xxx/Desktop/Example/cities.shp'
    cities = gpd.read_file('/Users/xxx/Desktop/Example/cities.shp')
    
    for item in cities.itertuples():
        print(item.gen)
        graph = ox.graph_from_polygon(item.geometry)
        ox.save_graphml(graph, filepath = '/Users/xxx/Desktop/'   item.gen   '.graphml')
    ---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
Input In [13], in <cell line: 1>()
      1 for item in cities.itertuples():
----> 2     print(item.gen)
      3     graph = ox.graph_from_polygon(item.geometry)
      4     ox.save_graphml(graph, filepath = '/Users/xxx/Desktop/'   item.gen   '.graphml')

AttributeError: 'Pandas' object has no attribute 'gen'

I tried to fix it with 'replace' or 'join' the false examples, but actually don't know how to incorporate it in the code. Does anyone has an idea?

CodePudding user response：

you have an extra = on the variable cities declaration.

Right now you have:

cities = gpd.read_file =('/Users/xxx/Desktop/Example/cities.shp')

That creates a variable cities, equal to a variable gpd.read_file, equal to a string with the value '/Users/xxx/Desktop/Example/cities.shp'

just remove the second equal so you call the function gpd.read_file with the string as parameter.

cities = gpd.read_file('/Users/xxx/Desktop/Example/cities.shp')

CodePudding user response：

as per comment, you are using itertuples() that returns named tuples. This means your code needs to consider situations where a name does not exist

name = item.gen if "gen" in item._fields else item.Index

have generated a full code MWE example of this. I really recommend using pathlib instead of hardcoded paths to files.

import geopandas as gpd
import osmnx as ox
from pathlib import Path

# use pathlib - better practice
p = Path.home().joinpath("Desktop/Examples")
if not p.exists():
    p.mkdir()

# path_to_folder = r'/Users/xxx/Desktop/Example/cities.shp'
# cities = gpd.read_file('/Users/xxx/Desktop/Example/cities.shp')
cities = gpd.read_file(gpd.datasets.get_path("naturalearth_cities"))

# need polygons - not really relevant to question
cities["NAME"] = cities["name"].str.upper()
cities_polys = gpd.read_file(
    "https://github.com/drei01/geojson-world-cities/raw/master/cities.geojson"
)
cities = (
    cities.drop(columns=["geometry"])
    .merge(cities_polys, on="NAME")
    .drop_duplicates(subset=["NAME"])
    .reset_index(drop=True)
    .head(5)
)
cities["gen"] = cities["NAME"]
# end of code not relevant to question

for item in cities.itertuples():
    name = item.gen if "gen" in item._fields else item.Index
    print(name)
    graph = ox.graph_from_polygon(item.geometry)
    ox.save_graphml(graph, filepath=p.joinpath(str(name)   ".graphml"))