I am doing a number of joins outside of ArcGIS as pandas seems to outperform it. I read in the dbf file with simpledbf, convert it to a dataframe, then do the join:

import pandas as pd
from simpledbf import Dbf5

#Source of the join

#Target of the join
dbf = Dbf5('C:\\Users\\myfolder\\mydbf.dbf')
d_b_f = dbf.to_dataframe()

d_b_f_2=pd.merge(d_b_f, data_source, on='COLUMN', how='left')

How can I save this dataframe as a dbf and overwrite the existing dbf?

I expected something along these lines to work:

final_dbf = Dbf5('C:\\Users\\myfolder\\finaldbf.dbf')
final_dbf = d_b_f_2.to_dbf()

but an error pops up saying that 'C:\Users\myfolder\finaldbf.dbf' does not exist. What do you suggest as a workaround? I wouldn't want to look into other packages, as I am sure there must be a way of doing this within simpledbf itself.


1) simpledbf cannot export dbf format as output (only CSV, SQL,pandas DataFrame, HDF5 Table)
2) Pandas DataFrames don't have a .to_dbf method

Simply use PySAL(dbf) and DBF files and the pandas DataFrame with the functions df2dbf(convert a pandas.DataFrame into a dbf), dbf2df(read a dbf file as a pandas.DataFrame) and appendcol2dbf(append a column and the associated data to a DBF) in dataIO.py) that you can adapt

import numpy as np
import pysal as ps
import pandas as pd
# adaptation of the function `df2dbf` to write your resulting dbf file
type2spec = {int: ('N', 20, 0),
             np.int64: ('N', 20, 0),
             float: ('N', 36, 15),
             np.float64: ('N', 36, 15),
             str: ('C', 14, 0),
             unicode: ('C', 14, 0)}
types = [type(d_b_f_2[i].iloc[0]) for i in d_b_f_2.columns]
specs = [type2spec[t] for t in types]
with ps.open("finaldbf.dbf", 'w') as db:
    db.header = list(df.columns)
    db.field_spec = specs
    for i, row in d_b_f_2.T.iteritems():
  • Interesting solution, but regardless if I try to overwrite the existing dbf or create a new one, I get the same error: AttributeError: __enter__ pointing at the line with ps.open(). Any advice would be much appreciated! – FaCoffee Jul 5 '18 at 15:10
  • look at line 47 of dataIO.py -> db = ps.open(dbf_path, 'w') – gene Jul 5 '18 at 15:22
  • Yes, I saw that. I also tried with 'w+' but got the same result. I don't get the meaning of that error message. – FaCoffee Jul 5 '18 at 15:25
  • What is your version of pysal ? – gene Jul 5 '18 at 16:09
  • My pysal version is 1.14.3 – FaCoffee Jul 6 '18 at 11:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.