In [1]:
import hkvfewspy as fewspy
import matplotlib.pyplot as plt
import os
import pandas as pd
In [2]:
fews_ts = fewspy.schemas.timeseries.FewsTimeSeries.from_pi_xml("./test_2_series.xml")
Plot the series by calling plot on the FewsTimeSeries object (plot is passed on to the plot function in pandas for Series and DataFrames).
In [3]:
fews_ts.plot(y="value")
plt.show()
View metadata:
In [4]:
fews_ts.header
Out[4]:
In [5]:
fews_ts.to_pi_xml("./temp_1series.xml")
In [6]:
fews_ts_collection = fewspy.schemas.timeseries.FewsTimeSeriesCollection.from_pi_xml("./test_2_series.xml")
Timeseries are collected in a DataFrame with header data as columns and the series in the events column:
In [7]:
fews_ts_collection.timeseries
Out[7]:
Loop through timeseries to plot
In [8]:
fig, ax = plt.subplots(1, 1)
for i in range(fews_ts_collection.timeseries.shape[0]):
series = fews_ts_collection.timeseries.events.iloc[i]
series.plot(y="value", ax=ax, label=series.header["stationName"])
plt.show()
Note that the objects stored in the timeseries DataFrame in the events column are actually FewsTimeSeries objects. This means that our earlier FewsTimeSeries object should be the same as the first timeseries in our FewsTimeSeriesCollection.
In [9]:
fews_ts_collection.timeseries.events[0]
Out[9]:
We will now test whether the FewsTimeSeries in the collection is the same as the single timeseries we imported earlier. But first, it is important to realize that when we wrote our earlier FewsTimeSeries object fews_ts to XML the header (endDate and startDate) were automatically updated from the data. We will do the same for the first timeseries in the FewsTimeSeriesCollection object.
In [10]:
# get start and end date from timeseries and update header
fews_ts_collection.timeseries.events[0]._update_header_dates()
Now test for equality:
In [11]:
fews_ts_collection.timeseries.events[0] == fews_ts
Out[11]:
It is possible to add data to the collection with the add_series method
In [12]:
help(fews_ts_collection.add_series)
Use data from first timeseries as "new" series:
In [13]:
new_series = fews_ts.timeseries
new_metadata = fews_ts.header
In [14]:
fews_ts_collection.add_series(new_series, new_metadata)
View our added timeseries in the DataFrame:
In [15]:
fews_ts_collection.timeseries
Out[15]:
We can write the FewsTimeSeriesCollection to file with the same command as before:
In [16]:
fews_ts_collection.to_pi_xml("./temp_3series.xml")
In [ ]:
# set client
pi = fewspy.pi
pi.setClient(wsdl='http://localhost:8081/FewsPiService/fewspiservice?wsdl')
In [ ]:
query = pi.setQueryParameters(prefill_defaults=True)
In [ ]:
query.parameterIds(['P.meting.dagcal'])
query.moduleInstanceIds(['ImportCAW'])
query.locationIds(['66011cal'])
query.startTime(pd.datetime(2010,1,1))
query.endTime(pd.datetime(2018,7,1))
query.clientTimeZone('Europe/Amsterdam')
query.query["onlyManualEdits"] = False
In [ ]:
query.query
In [ ]:
fews_ts2_coll = pi.getFewsTimeSeries(queryParameters=query)
In [ ]:
fews_ts2_coll
In [ ]:
fews_ts2_coll.timeseries
We queried specifically to get one timeseries, so extract FewsTimeSeries object from the collection:
In [ ]:
fews_ts2 = fews_ts2_coll.timeseries.events[0]
In [ ]:
fews_ts2.header
In [ ]:
fews_ts2.timeseries.head()
In [ ]:
fews_ts2.plot(y="value")
In [ ]: