Read, write and manipulate dfs0, dfs1, dfs2, dfs3, dfsu and mesh files.
Facilitates common data processing workflows for MIKE files.
For res1d and xns11 files use the related package MIKE IO 1D
- Windows or Linux operating system
- Python x64 3.6, 3.7,3.8 or 3.9
- (Windows) VC++ redistributables (already installed if you have MIKE)
- New ideas and feature requests - GitHub Discussions
- Bugs - GitHub Issues
- General help, FAQ - Stackoverflow with the tag
mikeio
From PyPI:
pip install mikeio
Or development version (main
is the default branch since 2021-04-23):
pip install https://github.com/DHI/mikeio/archive/main.zip
Generic read
method to read values, if you need additional features such as coordinates, use specialised classes instead e.g. Dfsu
>>> import mikeio
>>> ds = mikeio.read("random.dfs0")
>>> ds
<mikeio.Dataset>
Dimensions: (1000,)
Time: 2017-01-01 00:00:00 - 2017-07-28 03:00:00
Items:
0: VarFun01 <Water Level> (meter)
1: NotFun <Water Level> (meter)
>>> ds = mikeio.read("random.dfs1")
>>> ds
<mikeio.Dataset>
Dimensions: (100, 3)
Time: 2012-01-01 00:00:00 - 2012-01-01 00:19:48
Items:
0: testing water level <Water Level> (meter)
>>> from mikeio import Dfs0
>>> dfs = Dfs0('simple.dfs0')
>>> ts = dfs.to_dataframe()
>>> from datetime import datetime
>>> import numpy as np
>>> from mikeio import Dfs0
>>> data = [np.random.random([100])]
>>> dfs = Dfs0()
>>> dfs.write('simple.dfs0', data, start_time=datetime(2017, 1, 1), dt=60)
import pandas as pd
import mikeio
>>> df = pd.read_csv(
... "tests/testdata/co2-mm-mlo.csv",
... parse_dates=True,
... index_col="Date",
... na_values=-99.99,
... )
>>> df.to_dfs0("mauna_loa.dfs0")
For more examples on timeseries data see this notebook
>>> from mikeio import Dfs2
>>> dfs = Dfs2("random.dfs2")
>>> ds = dfs.read()
>>> ds
<mikeio.Dataset>
Dimensions: (3, 100, 2)
Time: 2012-01-01 00:00:00 - 2012-01-01 00:00:24
Items:
0: testing water level <Water Level> (meter)
For a complete example of conversion from netcdf to dfs2 see this notebook.
Another example of downloading meteorological forecast from the Global Forecasting System and converting it to a dfs2 ready to be used by a MIKE 21 model.
>>> import matplotlib.pyplot as plt
>>> from mikeio import Dfsu
>>> dfs = Dfsu("HD.dfsu")
>>> ds = dfs.read()
>>> idx = dfs.find_nearest_element(x=608000, y=6907000)
>>> plt.plot(ds.time, ds.data[0][:,idx])
>>> from mikeio import Mesh
>>> msh = Mesh("FakeLake.dfsu")
>>> msh.plot()
For more examples on working with dfsu and mesh see these notebooks:
- Basic dfsu
- 3d dfsu
- Mesh
- Speed & direction
- Dfsu and mesh plotting
- Export to netcdf
- Export to shapefile
Pfs is the format used for MIKE setup files (.m21fm, .m3fm, .sw etc.).
There is experimental support for reading pfs files, but the API is likely to change.
Useful when creating a new dfs file
>>> from mikeio.eum import EUMType, EUMUnit
>>> EUMType.Temperature
<EUMType.Temperature: 100006>
>>> EUMType.Temperature.units
[degree Celsius, degree Fahrenheit, degree Kelvin]
>>> EUMUnit.degree_Kelvin
degree Kelvin
MIKE IO is tested extensively. 95% total test coverage.
See detailed test coverage report below:
File Covered Missed %
-------------------------------------------------
mikeio\__init__.py 40 1 98%
mikeio\aggregator.py 103 9 91%
mikeio\bin\__init__.py 0 0 100%
mikeio\custom_exceptions.py 19 1 95%
mikeio\dataset.py 272 3 99%
mikeio\dfs.py 206 7 97%
mikeio\dfs0.py 239 15 94%
mikeio\dfs1.py 48 2 96%
mikeio\dfs2.py 100 2 98%
mikeio\dfs3.py 201 79 61%
mikeio\dfsu.py 1337 57 96%
mikeio\dfsutil.py 76 4 95%
mikeio\dotnet.py 63 4 94%
mikeio\eum.py 1230 3 99%
mikeio\generic.py 228 2 99%
mikeio\helpers.py 13 0 100%
mikeio\interpolation.py 54 1 98%
mikeio\spatial.py 279 4 99%
mikeio\xyz.py 12 0 100%
-------------------------------------------------
TOTAL 5082 229 95%
=================== 335 passed ==================
From MIKE IO v.0.7 it is now possible to run MIKE IO on Linux-based Cloud computing, e.g. Google Colab.