inazentra.py

from zentra.api import *
import os
from datetime import datetime, timedelta, date, timezone
import pandas as pd
import inafiles
from inaconf import inaconf
import matplotlib.dates as mdates
from matplotlib import pyplot



def login():

    logindata = inafiles.getZentraLogin()
    token = ZentraToken(username=logindata['username'][0], password=logindata['password'][0])
    return token

def read_data(token, device_sn, start_time, end_time):
    start = datetime.fromtimestamp(start_time)
    end = datetime.fromtimestamp(end_time)
    print('Requesting data from Zentra Cloud for %s, from %s - %s'%(device_sn, start, end))
    print('This might take a loooooong while... Zentra Cloud is not exactly a rocketship.')
    readings = ZentraReadings(sn=device_sn, token=token, start_time=start_time, end_time=end_time)
    return readings
# Report the readings from the first ZentraTimeseriesRecord

def read_precip(token, device_sn, start_time, end_time):
    # readings = read_data(token, device_sn,start_time,end_time)
    # data = pd.DataFrame(readings.timeseries[0].values)
    # precip_criterium = data['description'] == 'Precipitation'
    # precip_data = data[precip_criterium]
    # return precip_data
    return read_criterium(token, device_sn, start_time, end_time, 'Precipitation')
def read_batt(token, device_sn, start_time, end_time):
    return read_criterium(token, device_sn, start_time, end_time, 'Battery Percent')
def read_criterium(token, device_sn, start_time, end_time, criterium):
    readings = read_data(token, device_sn, start_time, end_time)
    data = pd.DataFrame(readings.timeseries[0].values)
    filter_criterium = data['description'] == criterium
    data = data[filter_criterium]
    return data

def saveplot(data, devicesn, filename):
    #grouped = data.groupby(data['dt'].dt.hour, as_index=False)['value'].sum().plot(kind='bar', x='datetime')
    #data.plot(x='datetime', y='value')
    #data.plot(x='datetime',y='rolling')
    #pyplot.show()
    end = datetime.now(timezone.utc)
    start = end - timedelta(days=30)
    lastmonth = (data['datetime'] > start) & (data['datetime'] <= end)
    data = data.loc[lastmonth]
    locationname = inafiles.getlocname(inafiles.getlocationfrommeteo(devicesn)[0])
    title ='%s (%s) 24h sum'%(locationname, devicesn)
    grouped = data.groupby(data['datetime'].dt.date)['value'].apply(lambda x: x.sum())
    grouped.plot(kind='bar', y='value', title=title)
    pyplot.savefig(filename)
    return filename
    #grouped2 = data.groupby(data['datetime'].dt.to_period('1h'),group_keys=True)['value'].apply(lambda x: x.sum())
    #grouped2 = data.groupby(data['datetime'].dt.time, group_keys=True)['value'].apply(lambda x: x.sum())

    #grouped2.plot(kind='bar', y='value', title='hourly sum')
    #pyplot.xticks(fontsize=6)
    #pyplot.gca().xaxis.set_major_locator(mdates.DayLocator())
    #pyplot.plot_date(grouped2)
    #pyplot.show()

def getmeteoforcam(camnr):
    nr = int(camnr)
    locdata = inafiles.getlocationdata(inaconf.get_camlocations(), nr)
    locnr = int(locdata[1])
    summary = pd.DataFrame(getsummary(locnr))
    print(summary)
    return summary

def getsummary(locationnr):
    nr = int(locationnr)
    enddate = datetime.now(timezone.utc)
    startdate = enddate - timedelta(days=inaconf.meteoreportdays)

    meteos = inafiles.getmeteofromlocation(nr)
    summary = []
    for device in meteos:
        data, file = readarchive(device)
        data['datetime']= pd.to_datetime(data['datetime'])
        relevant = data.loc[(data['datetime'] > startdate) & (data['datetime'] <= enddate)]
        total = relevant['value'].sum()
        maxval = relevant['value'].max()
        batt = gettodaysbattlevel(device)
        summary.append({'location': nr, 'device': device, 'start':startdate, 'end': enddate, 'total': total, 'max': maxval,'batt': batt})

    return summary

def gettodaysbattlevel(device_sn):
    token = login()
    end = int(datetime.now(timezone.utc).timestamp())
    start = int((datetime.now(timezone.utc) - timedelta(days=1)).timestamp())
    battdate = read_batt(token, device_sn, start_time=start,end_time=end)
    return battdate['value'].mean()



def readarchive(device_sn):
    maindir = inaconf.maindir
    archivefile = os.path.join(maindir, 'precip_%s.csv' % (device_sn))
    archivedata = inafiles.readDataframe(archivefile, ',')
    archivedata.columns = archivedata.columns.str.strip()
    if 'datetime' in archivedata.columns:
        archivedata['datetime'] = pd.to_datetime(archivedata['datetime'], utc=True)
    return archivedata, archivefile
def getmeteodata(locationnr):
    token = login()
    meteos = inafiles.getmeteofromlocation(locationnr)
    attachments = []
    for meteo in meteos:

        print('Reading Zentra archive...')
        archivedata, archivefile = readarchive(meteo)
        if 'datetime' in archivedata.columns:
            #archivedata['datetime']=pd.to_datetime(archivedata['datetime'], utc=True)
            first = archivedata['datetime'].max()
        else:
            first = datetime(year=2020, month=8,  day=27)
        last = datetime.now()


        data = read_precip(token=token, device_sn=meteo, start_time=int(first.timestamp()), end_time=int(last.timestamp()))

        all = pd.concat([data, archivedata],ignore_index=True)
        all['datetime']=pd.to_datetime(all['datetime'], utc=True)
        all.set_index('datetime')
        all['rolling']=all['value'].rolling(min_periods=1, window=48).sum()
        all=all.drop_duplicates(subset=['mrid', 'description']).reset_index(drop=True)
        all = all[['datetime','mrid','rssi','port','units','description','value','error']]
        all.to_csv(archivefile)
        data = all
        #data['date'] = data['datetime']
        file = os.path.join(inaconf.maindir, 'plot_%s.png'%(meteo))
        plot = saveplot(data, devicesn=meteo, filename = file)
        attachments.append({'device': meteo, 'plotfile': plot})
    return attachments