mikeio 0.3.0

A package that works with the DHI dfs libraries to facilitate creating, writing and reading dfs0, dfs2, dfs3, dfsu and mesh files.

mikeio: input/output of MIKE files in python

Facilitate creating, reading and writing dfs0, dfs2, dfs1 and dfs3, dfsu and mesh files. Reading Res1D data.

Requirements

• Python x64 >=3.6
• MIKE or MIKE SDK

Installation

From PyPI:

pip install mikeio

Or development version:

pip install https://github.com/DHI/mikeio/archive/master.zip

Examples

Reading data from dfs0, dfs1, dfs2, dfsu

>>> import mikeio
>>> ds = mikeio.read("random.dfs0")
>>> ds
DataSet(data, time, names)
Number of items: 2
Shape: (1000,)
2017-01-01 00:00:00 - 2017-07-28 03:00:00

>>> ds = mikeio.read("random.dfs1")
>>> ds
DataSet(data, time, names)
Number of items: 1
Shape: (100, 3)
2012-01-01 00:00:00 - 2012-01-01 00:19:48

Reading dfs0 file into Pandas DataFrame

from mikeio import Dfs0
dfs = Dfs0()
ts = dfs.read_to_pandas('simple.dfs0')

Create simple timeseries

from datetime import datetime, timedelta
import numpy as np
from mikeio import Dfs0

# create a list containing data for each item
data = []

# Some random values for first (and only) item
d = np.random.random([100])
data.append(d)

dfs = Dfs0()
dfs.create(filename='simple.dfs0',
           data=data,
           start_time=datetime(2017, 1, 1),
           dt=60)

Create equidistant dfs0 with weekly timestep

from mikeio import Dfs0
from mikeio.eum import TimeStep
d1 = np.random.random([1000])
d2 = np.random.random([1000])
data = []
data.append(d1)
data.append(d2)

dfs = Dfs0()
dfs.create(filename='random.dfs0',
           data=data,
           start_time=datetime(2017, 1, 1),
           timeseries_unit=TimeStep.DAY,
           dt=7,
           names=['Random1', 'Random2'],
           title='Hello Test')

For more examples on timeseries data see this notebook

Read dfs2 data

from mikeio import Dfs2
dfs2File = r"20150101-DMI-L4UHfnd-NSEABALTIC-v01-fv01-DMI_OI.dfs2"
dfs = Dfs2()
res = dfs.read(dfs2File)
res.names

Create dfs2

For a complete example of conversion from netcdf to dfs2 see this notebook.

Another example of downloading meteorlogical forecast from the Global Forecasting System and converting it to a dfs2 ready to be used by a MIKE 21 model.

Read Res1D file Return Pandas DataFrame

import res1d as r1d
p1 = r1d.ExtractionPoint()
p1.BranchName  = 'branch1'
p1.Chainage = 10.11
p1.VariableType  = 'Discharge'
ts = r1d.read('res1dfile.res1d', [p1])

Read dfsu files

import matplotlib.pyplot as plt
from mikeio import Dfsu

dfs = Dfsu()

filename = "HD.dfsu"
res = dfs.read(filename)

idx = dfs.find_closest_element_index(x=608000, y=6907000)

# data has two dimensions time, x
plt.plot(res.time, res.data[0][:,idx])

Misc utilities

to query variable type, time series types (useful when creating a new dfs file)

>>> from mikeio.dfs_util import type_list, unit_list
>>> type_list('Water level')
{100000: 'Water Level', 100307: 'Water level change'}

>>> unit_list(100307)
{1000: 'meter', 1003: 'feet'}