Get Series Data

The cli.getSeries() series function is without a doubt the most important package function. It is the core function to get data out of the BayEOS-Server. An active connection is necessary to call the function. Let's open a connection to our BayEOS Server by reusing the guest alias of our CLI Example.



In [1]:

    
from bayeos import cli as bayeos
bayeos.connect('guest@bayeos')









    



Open connection to http://bayeos.bayceer.uni-bayreuth.de/BayEOS-Server/XMLServlet as user gast






    Out[1]:





True

Default Parameters

A first look at the function definition showed, that there are several ways to call the function. The easiest approach is to navigate into a folder with data series and call the getSeries() function with it's default arguments. This will produce a numpy array with channel columns and rows of data from yesterday until todoay for all series in your current folder.



In [2]:

    
bayeos.cd(36076)









    



Alle Messungen/Micrometeorology Dept/AWS Ecological Botanical Garden/AWS EBG Data start May 08, 1997/Lufttemperatur & Feuchte
14294     |..
2718      |Lufttemperatur 02 m Höhe (HMP45)                  |messung_massendaten |1997-05-08 06:00:00 |2014-08-26 14:30:00 
2717      |Relative Luftfeuchte 02 m Höhe (HMP45)            |messung_massendaten |1997-05-08 06:00:00 |2014-08-26 14:30:00



In [3]:

    
data = bayeos.getSeries()









    



Fetching data of series:[2718, 2717]
Rows fetched:232

Let's check the type of our data variable.



In [4]:

    
type(data)









    Out[4]:





numpy.ndarray

Each numpy array contains metainformation in a property called dtype. It's up to you to reuse these properties during the analysis of your data.



In [5]:

    
data.dtype









    Out[5]:





dtype([('dateTime', '>i4'), ('Lufttemperatur 02 m Höhe (HMP45)', '>f4'), ('Relative Luftfeuchte 02 m Höhe (HMP45)', '>f4')])

The first column of our data array contains values as integer. These values represent the time of observation in seconds since January 1, 1970, 00:00:00 GMT. Further columns represent the series values as Floats. You can produce python dateTime typed values out of the first column values with the following statement:



In [6]:

    
import datetime as dt
dateCol = [dt.datetime.fromtimestamp(t) for t in data['dateTime']]

Select of Series by ID

Series can be selected by id or a list of ids.



In [7]:

    
data = bayeos.getSeries(2718)









    



Rows fetched:232



In [8]:

    
data = bayeos.getSeries([2718,2717])









    



Rows fetched:232

Filter Rows by Start and/or Until Date of Observation

You can filter the series values by a start and/or until value. Both parameters must fullfill the following restrictions:

start := {'yesterday'|'today'|'%Y-%m-%d %H:%M'}
until := {'yesterday'|'today'|'%Y-%m-%d %H:%M'}

Here are examples showing the different possibilities to call the function:



In [9]:

    
data = bayeos.getSeries([2718,2717],start='yesterday')









    



Rows fetched:232



In [10]:

    
data = bayeos.getSeries([2718,2717],start='yesterday',until='today')









    



Rows fetched:232



In [11]:

    
date = bayeos.getSeries([2718,2717],start='2014-08-01 00:00',until='2014-08-07 00:00')









    



Rows fetched:865

Filter Rows by Time Interval of Observation

There are several fixed time intervals defined to pick a time period:

today
this week
this month
this year
yesterday
last week
last month
last year

You can use the interval name and ommit the start end until parameter to filter the data:



In [12]:

    
data = bayeos.getSeries([2718,2717],interval="last week")









    



Rows fetched:1009

Select Aggregated Values

Server side aggregated series values can be accessed by calling the getSeries() function with aggFunc and aggInt parameters:



In [13]:

    
data = bayeos.getSeries([2718,2717],interval="last week",aggfunc="avg",aggint="day")









    



Rows fetched:8

The available functions can be determined at runtime:



In [14]:

    
print("Aggregation functions:" + str([x for x in bayeos.getAggFunc()]))
print("Aggregation intervals:" + str([x for x in bayeos.getAggInt()]))









    



Aggregation functions:['avg', 'max', 'sum', 'min']
Aggregation intervals:['30min', 'day', 'hour', 'month', 'year']

CSFlags

If set for aggegarated values, additional count columns will be added to the output array.



In [15]:

    
data = bayeos.getSeries([2718,2717],interval="last week",aggfunc="avg",aggint="day",csFlag=True)









    



Rows fetched:8

If set for non aggregated values, status flag column (range(10)) will be added for each value column.



In [16]:

    
data = bayeos.getSeries([2718,2717],interval="last week",csFlag=True)









    



Rows fetched:1009



In [17]:

    
data.dtype









    Out[17]:





dtype([('dateTime', '>i4'), ('Lufttemperatur 02 m Höhe (HMP45)', '>f4'), ('Lufttemperatur 02 m Höhe (HMP45)_status', '>i4'), ('Relative Luftfeuchte 02 m Höhe (HMP45)', '>f4'), ('Relative Luftfeuchte 02 m Höhe (HMP45)_status', '>i4')])