c-h-david/rrr

Reproducible Routing Rituals (RRR)

rrr's Issues

rrr_cat_tot_gen_one_nhdplus.py DbfError

I have installed rrr and rapid with docker (on a Windows 10 computer) and am trying to prepare input files. Specifically, I'm trying to prepare the csv file from an NHDPlus catchment file using rrr_cat_tot_gen_one_nhdplus.py. I copied the files from my local computer to the docker container using the following commands:

# HUC10_catch = catchment shapefile
docker cp HUC10_catch.dbf <container ID>:/home/rrr/src/HUC10_catch.dbf
docker cp HUC10_catch.shp <container ID>:/home/rrr/src/HUC10_catch.shp
docker cp HUC10_catch.shx <container ID>:/home/rrr/src/HUC10_catch.shx

Next, I tried running the the rrr_cat_tot_gen_one_nhdplus python script using the following command:

docker exec <container ID> python /home/rrr/src/rrr_cat_tot_gen_one_nhdplus.py /home/rrr/src/HUC10_catch.dbf /home/rrr/src/rapid_catchment_HUC10.csv

And the following error is thrown:

Traceback (most recent call last):
  File "/home/rrr/src/rrr_cat_tot_gen_one_nhdplus.py", line 74, in <module>
    nhd_cat_dbf=dbf.Table(nhd_cat_file)
  File "/usr/local/lib/python2.7/dist-packages/dbf/ver_2.py", line 4870, in __new__
Command line inputs
- /home/rrr/src/HUC10_catch.dbf
- /home/rrr/src/rapid_catchment_HUC10.csv
Read shapefile
- Read attributes
    possibles = guess_table_type(filename)
  File "/usr/local/lib/python2.7/dist-packages/dbf/ver_2.py", line 8052, in guess_table_type
    reported = table_type(filename)
  File "/usr/local/lib/python2.7/dist-packages/dbf/ver_2.py", line 8081, in table_type
    raise DbfError("Unknown dbf type: %s (%x)" % (version, ord(version)))
dbf.ver_2.DbfError: Unknown dbf type:   (0)

I looked at the table_type documentation and the version_map documentation and discovered that it seems to not be recognizing the dbf version correctly, therefore throwing the DbfError. In order to debug locally, I created a conda environment on my computer to run the dbf.Table(nhd_cat_file) command, which did not throw any errors. In addition, when I run the following code in the conda environment, it seems that the dbf file version matches one of the versions in the version_map documentation.

>> fd = open('HUC10_catch.dbf')
>> version = fd.read(1)
>> version
'\x03'

Any idea why the issue seems to only exist in the docker container? I am new to using docker, so please let me know if I'm doing something wrong.

Thanks!

Change legend in plots to reflect the model name used in the header of CSV files

@kandread, would you be able to implement this modification?