"""
This file contains the import routine for import of groundwater observations
from RWS Waterbase / WaterInfo database. (http://waterinfo.rws.nl/)
Author: R.A. Collenteur, Artesia Water 2017
"""
from pandas import read_csv
from ..timeseries import TimeSeries
[docs]def read_waterbase(fname, locations=None, variable="NUMERIEKEWAARDE",
kind="waterlevel", freq="10min"):
"""Method to import waterlevel ts from waterbase.
Parameters
----------
fname: str
string with the path and filename of the waterbase file.
variable: str
name of the variable to collect the time series from. Only one
variable name is allowed.
kind: str
freq: str
Returns
-------
ts: pastas.TimeSeries
returns a Pastas TimeSeries object or a list of objects.
Notes
-----
More information on the ts provided by the Waterbase database see:
http://waterinfo.rws.nl/
the xy-coordinates are calculates as the mean xy-coordinate in case these
values are not unique.
"""
ts = []
df = read_csv(fname, delimiter=";", index_col="Date", decimal=",",
usecols=["MEETPUNT_IDENTIFICATIE", "WAARNEMINGDATUM",
"WAARNEMINGTIJD", variable, "EPSG", "X", "Y"],
parse_dates={"Date": ["WAARNEMINGDATUM", "WAARNEMINGTIJD"]},
infer_datetime_format=True, dayfirst=True,
na_values=[-999999999, 999999999],
encoding="ISO-8859-1")
if locations is None:
locations = df.MEETPUNT_IDENTIFICATIE.unique()
elif isinstance(locations, str):
locations = [locations]
for name in locations:
series = df.loc[df["MEETPUNT_IDENTIFICATIE"].isin([name])]
metadata = {
"x": series.X.mean(),
"y": series.Y.mean(),
"z": 0,
"projection": "epsg:" + str(series.loc[:, "EPSG"].unique()[0]),
"units": "cm"
}
series = series.loc[:, variable].sort_index()
ts.append(TimeSeries(series, name=name, metadata=metadata,
settings=kind, freq_original=freq))
if len(ts) == 1:
ts = ts[0]
return ts