generic

generic

Generic functions for working with all types of dfs files.

Functions

Name	Description
add	Add two dfs files (a+b).
avg_time	Create a temporally averaged dfs file.
change_datatype	Change datatype of a DFS file.
concat	Concatenates files along the time axis.
diff	Calculate difference between two dfs files (a-b).
extract	Extract timesteps and/or items to a new dfs file.
fill_corrupt	Replace corrupt (unreadable) data with fill_value, default delete value.
quantile	Create temporal quantiles of all items in dfs file.
scale	Apply scaling to any dfs file.
transform	Transform a dfs file by applying functions to items.

add

generic.add(infilename_a, infilename_b, outfilename)

Add two dfs files (a+b).

avg_time

generic.avg_time(infilename, outfilename, skipna=True)

Create a temporally averaged dfs file.

Parameters

Name	Type	Description	Default
infilename	str | pathlib.Path	input filename	required
outfilename	str | pathlib.Path	output filename	required
skipna	bool	exclude NaN/delete values when computing the result, default True	True

change_datatype

generic.change_datatype(infilename, outfilename, datatype)

Change datatype of a DFS file.

The data type tag is used to classify the file within a specific modeling context, such as MIKE 21. There is no global standard for these tags—they are interpreted locally within a model setup.

Application developers can use these tags to classify files such as bathymetries, input data, or result files according to their own conventions.

Default data type values assigned by MikeIO when creating new files are: - dfs0: datatype=1 - dfs1-3: datatype=0 - dfsu: datatype=2001

Parameters

Name	Type	Description	Default
infilename	str | pathlib.Path	input filename	required
outfilename	str | pathlib.Path	output filename	required
datatype	int	DataType to be used for the output file	required

Examples

>>> change_datatype("in.dfsu", "out.dfsu", datatype=107)

concat

generic.concat(infilenames, outfilename, keep='last')

Concatenates files along the time axis.

Overlap handling is defined by the keep argument, by default the last one will be used.

Parameters

Name	Type	Description	Default
infilenames	Sequence[str | pathlib.Path]	filenames to concatenate	required
outfilename	str | pathlib.Path	filename of output	required
keep	str	either ‘first’ (keep older), ‘last’ (keep newer) or ‘average’ can be selected. By default ‘last’	'last'

Notes

The list of input files have to be sorted, i.e. in chronological order

diff

generic.diff(infilename_a, infilename_b, outfilename)

Calculate difference between two dfs files (a-b).

Parameters

Name	Type	Description	Default
infilename_a	str | pathlib.Path	full path to the first input file	required
infilename_b	str | pathlib.Path	full path to the second input file	required
outfilename	str | pathlib.Path	full path to the output file	required

Examples

from mikeio import generic
generic.diff("../data/oresundHD_run1.dfsu", "../data/oresundHD_run2.dfsu", "diff.dfsu");

  0%|          | 0/5 [00:00<?, ?it/s]100%|██████████| 5/5 [00:00<00:00, 2645.25it/s]

extract

generic.extract(infilename, outfilename, start=0, end=-1, step=1, items=None)

Extract timesteps and/or items to a new dfs file.

Parameters

Name	Type	Description	Default
infilename	str | pathlib.Path	path to input dfs file	required
outfilename	str | pathlib.Path	path to output dfs file	required
start	(int, float, str or datetime)	start of extraction as either step, relative seconds or datetime/str, by default 0 (start of file)	0
end	(int, float, str or datetime)	end of extraction as either step, relative seconds or datetime/str, by default -1 (end of file)	-1
step	int	jump this many step, by default 1 (every step between start and end)	1
items	(int, list(int), str, list(str))	items to be extracted to new file	None

Examples

>>> extract('f_in.dfs0', 'f_out.dfs0', start='2018-1-1')
>>> extract('f_in.dfs2', 'f_out.dfs2', end=-3)
>>> extract('f_in.dfsu', 'f_out.dfsu', start=1800.0, end=3600.0)
>>> extract('f_hourly.dfsu', 'f_daily.dfsu', step=24)
>>> extract('f_in.dfsu', 'f_out.dfsu', items=[2, 0])
>>> extract('f_in.dfsu', 'f_out.dfsu', items="Salinity")
>>> extract('f_in.dfsu', 'f_out.dfsu', end='2018-2-1 00:00', items="Salinity")

fill_corrupt

generic.fill_corrupt(infilename, outfilename, fill_value=np.nan, items=None)

Replace corrupt (unreadable) data with fill_value, default delete value.

Parameters

Name	Type	Description	Default
infilename	str | pathlib.Path	full path to the input file	required
outfilename	str | pathlib.Path	full path to the output file	required
fill_value	float	value to use where data is corrupt, default delete value	np.nan
items	Sequence[str | int] | None	Process only selected items, by number (0-based) or name, by default: all	None

quantile

generic.quantile(
    infilename,
    outfilename,
    q,
    *,
    items=None,
    skipna=True,
    buffer_size=1000000000.0,
)

Create temporal quantiles of all items in dfs file.

Parameters

Name	Type	Description	Default
infilename	str | pathlib.Path	input filename	required
outfilename	str | pathlib.Path	output filename	required
q	float | Sequence[float]	Quantile or sequence of quantiles to compute, which must be between 0 and 1 inclusive.	required
items	Sequence[int | str] | int | str | None	Process only selected items, by number (0-based) or name, by default: all	None
skipna	bool	exclude NaN/delete values when computing the result, default True	True
buffer_size	float	for huge files the quantiles need to be calculated for chunks of elements. buffer_size gives the maximum amount of memory available for the computation in bytes, by default 1e9 (=1GB)	1000000000.0

Examples

>>> quantile("in.dfsu", "IQR.dfsu", q=[0.25,0.75])

>>> quantile("huge.dfsu", "Q01.dfsu", q=0.1, buffer_size=5.0e9)

>>> quantile("with_nans.dfsu", "Q05.dfsu", q=0.5, skipna=False)

scale

generic.scale(infilename, outfilename, offset=0.0, factor=1.0, items=None)

Apply scaling to any dfs file.

Parameters

Name	Type	Description	Default
infilename	str | pathlib.Path	full path to the input file	required
outfilename	str | pathlib.Path	full path to the output file	required
offset	float	value to add to all items, default 0.0	0.0
factor	float	value to multiply to all items, default 1.0	1.0
items	Sequence[int | str] | None	Process only selected items, by number (0-based) or name, by default: all	None

transform

generic.transform(infilename, outfilename, vars, keep_existing_items=True)

Transform a dfs file by applying functions to items.

Parameters

Name	Type	Description	Default
infilename	str | pathlib.Path	full path to the input file	required
outfilename	str | pathlib.Path	full path to the output file	required
vars	Sequence[DerivedItem]	List of derived items to compute.	required
keep_existing_items	bool	If True, existing items in the input file will be kept in the output file. If False, only the derived items will be written to the output file. Default is True.	True

Examples

import numpy as np
import mikeio
from mikeio.generic import DerivedItem, transform
item = DerivedItem(
        name="Current Speed",
        type=mikeio.EUMType.Current_Speed,
        func=lambda x: np.sqrt(x["U velocity"] ** 2 + x["V velocity"] ** 2),
    )
transform(
    infilename="../data/oresundHD_run1.dfsu",
    outfilename="out.dfsu",
    vars=[item],
    keep_existing_items=False,
)