visum_to_pandas's Introduction

visum_to_pandas

python scripts to parse visum .net and .dmd file to pandas and store as .csv files (c) Rafal Kucharski, Politechnika Krakowska, Krakow, Poland

usage

open ptv_visum_to_pandas.py and go to main() where you can find the below pipeline:

0. Set your paths - your ver files and export paths

NETPATH = "./test/MOMM_net.net"  # .net file to parse into csv tables
DMDPATH = "./test/MOMM_full_dmd.dmd"  # .dmd file to parse into csv tables
OUTPATH = "./test/data/"  # save the .csv's here
VERPATH = "./test/test_matrices.ver"

1. Export .ver file to editable .net and .dmd

export_net_dmd(path=VERPATH)

2. export network to csv

parse(NETPATH, export_path=OUTPATH) # a) full export
parse(NETPATH, export_path=OUTPATH, export_list=["Links", "Nodes"])  # b) just some tables

3. export demand files to csv

parse(DMDPATH, export_path=OUTPATH)

4. Export matrices (export from .dmd is not very useful)

matrices_export_via_com(VERPATH, export_path=OUTPATH)  # a) full export
matrices_export_via_com(VERPATH, export_path=OUTPATH, export_list=[101, 102])   # b) just some matrices

5. see if everything went fine and understand how to use it further with pandas

test_read("./test/data/Tripgeneration.csv")
test_read("./test/data/Links.csv")
test_read("./test/data/Mtx_10.csv")

Third party

We link PTV Visum with pandas

test

run test.py to see if the links.csv is as we expect and if matrix entries.csv are the same. Use only for the original files.

contributors

https://github.com/PatrikHlobil

possible modifications

pd.DataFrame(data, columns=cols).to_csv(OUTPATH+table_name+".csv") can be easily modified to be used directly in pandas (e.g. dict of DataFrames, or h5 store).
~~pipeline can be smoother if this is called from within Visum, you shall add Visum.IO.SaveNet, Visum.IO.SaveDmd to prepare the files under good paths and then use them in parse~~ done

visum_to_pandas's People

Contributors

Stargazers

Watchers

visum_to_pandas's Issues

Improvement of parse function

Hi Rafal,

I have to say that your visum_to_pandas script is great. I worked a bit on the most recent version and made the following changes/improvement for the "parse" function:

use os.path.join for the combination of paths
use "string".split(9 to get rid of newline and space characters
auto-detect COL_SEPARATOR
do not export index of DataFrame
add parameters (export_path, export) to the parse function:

:param export_path (optional): Path to folder, where to save the CSV files.
If not specified, the CSV files are save in
folder of the net files.
:param export (optional): List of network objects that should be exported. If
not specified, all network objects will be exported.

If you like, you can just include my changes in your code. The file was uploaded with the name "ptv_visum_to_pandas_2.py".

Best regards,

Patrik

Usability and capsulation