PFS

A configuration file format used in MIKE software

A PFS file is a text file with a tree structure that contains parameters and settings for MIKE tools and engines. MIKE IO can read, modify and create PFS files.

The PFS file

The content of the PFS file is similar to a nested dictionary. The root element is often called the target. Some PFS files have multiple root elements. The below sections are called PFS Sections which can be nested and contain key-value pairs called keywords and parameters.

[TARGET1]
   keywordA = parameterA
   [SECTION1]
      keywordB = parameterB
      keywordC = parameterC
      [SECTION2]
         keywordD = parameterD
      EndSect  // SECTION2
   EndSect  // SECTION1
EndSect  // TARGET1

[TARGET2]
   keywordE = parameterE
   [SECTION3]
      keywordF = parameterF
   EndSect  // SECTION3
EndSect  // TARGET2

1: First target, access this part with pfs.targets[0]
2: Second target, access this part with pfs.targets[1]

Note

Comments // is used to add comments to the PFS file (e.g. // SECTION2), the comments are ignored by MIKE IO.

Read

When a PFS file is read with MIKE IO, a PfsDocument object is created. It will contain one or more PfsSection objects - one for each target. The PfsSections will typically contain other PfsSections together with a number of key-value pairs.

A PFS file is read using mikeio.read_pfs:

import mikeio

pfs = mikeio.read_pfs("../data/pfs/concat.mzt")
pfs

[txconc]
   CLSID = 'TxConc.dll'
   TypeName = 'txconc'
   CREATEDTIME = '2020-03-11T15:24:45'
   MODIFIEDTIME = '2020-03-11T15:24:45'
   NOTES = ''
   [Setup]
      Name = 'Setup Name'
      NumberFiles = 2
      NumberDimensions = 1
      NumberItems = 1
      InsertDelIfGabs = 0
      UseTimeRange = 1
      TimeRange = 0, 145, 1
      FirstFileTimeDef = 0
      OverwriteWithLatest = 1
      [File_1]
         InputFile = |.\tide1.dfs1|
         Items = 1
      EndSect  // File_1
      [File_2]
         InputFile = |.\tide2.dfs1|
         Items = 1
      EndSect  // File_2
      [File_Out]
         OutputFile = |.\txconc.dfs1|
         OutputFileTitle = ''
      EndSect  // File_Out
   EndSect  // Setup
EndSect  // txconc

PfsDocument

The mikeio.PfsDocument is the MIKE IO equivalent to a PFS file. Its targets can be accessed by their name (as properties), like this:

pfs.txconc

CLSID = 'TxConc.dll'
TypeName = 'txconc'
CREATEDTIME = '2020-03-11T15:24:45'
MODIFIEDTIME = '2020-03-11T15:24:45'
NOTES = ''
[Setup]
   Name = 'Setup Name'
   NumberFiles = 2
   NumberDimensions = 1
   NumberItems = 1
   InsertDelIfGabs = 0
   UseTimeRange = 1
   TimeRange = 0, 145, 1
   FirstFileTimeDef = 0
   OverwriteWithLatest = 1
   [File_1]
      InputFile = |.\tide1.dfs1|
      Items = 1
   EndSect  // File_1
   [File_2]
      InputFile = |.\tide2.dfs1|
      Items = 1
   EndSect  // File_2
   [File_Out]
      OutputFile = |.\txconc.dfs1|
      OutputFileTitle = ''
   EndSect  // File_Out
EndSect  // Setup

Or by the pfs.targets object (which is a list of PfsSections). Each of the targets is a PfsSection object consisting of key-value pairs (keyword-parameter) and other PfsSections.

The mikeio.PfsDocument object is similar to a dictionary. You can loop over its contents with items(), keys() and values() like a dictionary.

PfsSection

The mikeio.PfsSection object is also similar to a dictionary. You can loop over its contents with items(), keys() and values() like a dictionary.

pfs.txconc.keys()

dict_keys(['CLSID', 'TypeName', 'CREATEDTIME', 'MODIFIEDTIME', 'NOTES', 'Setup'])

You can access a specific parameter with the get() method:

pfs.txconc.get("CLSID")

'TxConc.dll'

Or as a property with dot-notation—which is prefered in most cases as it is more readable:

pfs.txconc.CLSID

'TxConc.dll'

A PfsSection can be converted to a dictionary with the to_dict() method:

pfs.txconc.Setup.File_1.to_dict()

{'InputFile': '|.\\tide1.dfs1|', 'Items': 1}

If a PfsSection contains enumerated subsections, they can be converted to a pandas DataFrame with the to_dataframe() method:

pfs.txconc.Setup.to_dataframe(prefix="File_")

	InputFile	Items
1	\|.\tide1.dfs1\|	1
2	\|.\tide2.dfs1\|	1

Unique or non-unique keywords

Depending on the engine intended for reading the PFS file it may or may not make sense to have multiple identical keywords in the same PfsSection. MIKE 21/3 and the marine tools does not support non-unique keywords—if non-unique keywords are present, only the first will be read and the presence is most likely a mistake made by hand-editing the file. In other tools, e.g. MIKE Plot Composer, non-unique keywords are used a lot. How MIKE IO shall deal with non-unique keywords can be specified using the unique_keywords argument in the mikeio.read_pfs function:

pfs = mikeio.read_pfs("myplot.plt", unique_keywords=False)

If a PfsSection contains non-unique PfsSections or keywords and unique_keywords=False, the repeated key will only appear once and the corresponding value will be a list.

Update

The PfsSection object can be modified. Existing values can be changed, new key-value pairs can be added, subsections can added or removed.

Modify existing keyword

It is very simple to modify an existing keyword:

pfs.txconc.Setup.Name = "new name"

Add new key-value pair

A new key-value pair can be added, like in a dictionary, in this way:

pfs.txconc.Setup["NewKeyword"] = 12.0

Add new section as a copy of another section

Often a PfsSection is added using an existing PfsSection as a template.

s = pfs.txconc.Setup.File_1.copy()
s.InputFile = '|.\tide3.dfs1|'
pfs.txconc.Setup["File_3"] = s

Add new section from a dictionary

A PfsSection can be created from a dictionary and then added to another PfsSection like any other key-value pair:

d = {'InputFile': '|.\\tide4.dfs1|', 'Items': 1}
s = mikeio.PfsSection(d)
pfs.txconc.Setup["File_4"] = s

Write to file

A Pfs document can be written to a file using the write method.

pfs.write("new.pfs")

Create new Pfs files

A new PFS file can be created from dictionary in the following way:

d = dict(
    key1=1,
    lst=[0.3, 0.7],
    file_name=r"|path\file.dfs0|",
    start_time=[2019, 7, 1, 0, 0, 0],
)
pfs = mikeio.PfsDocument({"MYTOOL": d})
pfs

[MYTOOL]
   key1 = 1
   lst = 0.3, 0.7
   file_name = |path\file.dfs0|
   start_time = 2019, 7, 1, 0, 0, 0
EndSect  // MYTOOL

Multiple targets can be achieved by providing list of dictionaries, in this way you can create a PFS file with multiple targets for the same tool.

t1 = {"file_name": r"|path\file1.dfs0|"}
t2 = {"file_name": r"|path\file2.dfs0|"}

pfs = mikeio.PfsDocument([t1, t2], names=["ATOOL", "ATOOL"])
pfs

[ATOOL]
   file_name = |path\file1.dfs0|
EndSect  // ATOOL
[ATOOL]
   file_name = |path\file2.dfs0|
EndSect  // ATOOL