GeoPandas – Fixing AttributeError: ‘DataFrame’ Object Has No Attribute ‘to_file’

geodataframegeopandas

I have turned some R scripts for data cleaning/standardization of shapefiles into Python scripts.

Here is the R code:

x<-c(farm,field,"trialyield",2017)
file<-paste(x, collapse="_")
yield <- readOGR(".", file)
#identify column of dry yield
yield.df<-data.table::as.data.table(yield)
yield<-yield[,4]
names(yield@data)[1] <- "yield"

and here is what I did to try and convert the block of R code in Python:

y = farm + ' ' + field + ' ' + 'trialyield 2017'
y = y.replace(' ', '_')

trialyield = gpd.read_file(y)

title_col = trialyield.columns[0]
name_map = dict(zip(trialyield.columns[[4]], ['yield']))
trialyield.rename(columns=name_map, inplace=True)
trialyield[[title_col, 'yield']].to_file('trialyield_output.shp')

However, I get the current attribute error:

Traceback (most recent call last):
  File "./sommer_uofiaddington1_cleanandagg_2017.py", line 48, in <module>
    trialyield[[title_col, 'yield']].to_file('trialyield_output.shp')
  File "/opt/anaconda3/lib/python3.6/site-packages/pandas/core/generic.py", 
line 3614, in __getattr__
    return object.__getattribute__(self, name)
AttributeError: 'DataFrame' object has no attribute 'to_file'

if I recall gpd.read_file returns a GeodataFrame.

Why is it returning a DataFrame?

Best Answer

If you don't select geometry column from a GeoDataFrame, you get a DataFrame.

For example:

print type(trialyield[['column1', 'column2']])
# OUT:
# pandas.core.frame.DataFrame

print type(trialyield[['column1', 'column2', 'geometry']])
# OUT:
# geopandas.geodataframe.GeoDataFrame

Change the last line in following way:

trialyield[[title_col, 'yield', 'geometry']].to_file('trialyield_output.shp')
Related Question