In [4]:

    

import requests
import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline


    

In [115]:

    

r = requests.get("http://glenecho.stream:8080/list").json()


    

In [116]:

    

humidity = [i for i in r if i['measurement'] == 'humidity']
absolute_humidity = [i for i in r if i['measurement'] == 'absolute_humidity']
temp = [i for i in r if i['measurement'] == 'temperature' and i['sensor'] == 'bme680']
pressure = [i for i in r if i['measurement'] == 'pressure']
air = [i for i in r if i['measurement'] == 'gas_resistance']
depth = [i for i in r if i['measurement'] == 'depth']


    

In [117]:

    

humidity_value = np.array([i['value'] for i in humidity])
humidity_time = np.array([i['timestamp'] for i in humidity])
temp = np.array([i['value'] for i in temp])
pressure = np.array([i['value'] for i in pressure])
air_value = np.array([i['value'] for i in air])
air_time = np.array([i['timestamp'] for i in air])
depth = np.array([i['value'] for i in depth])
absolute_humidity = np.array([i['value'] for i in absolute_humidity])


    

In [57]:

    

plt.scatter(absolute_humidity[500:500+50], np.log(air_value[500:500+50]))
plt.xlim([5, 14])
plt.ylim([10.5, 13.5])


    

    
Out[57]:





(10.5, 13.5)






    

In [25]:

    

import scipy.stats


    

In [60]:

    

for i in range(40):
    offset = i * 25
    regress = scipy.stats.linregress(absolute_humidity[offset:offset+50], np.log(air_value[offset:offset+50]))
    if (regress.rvalue**2 > 0.6):
        print(offset)
        print(regress.slope, regress.intercept, regress.rvalue**2)


    

    




0
(-0.8004073838025771, 19.79319301037074, 0.8050542157575973)
100
(-0.23112912443394548, 13.732953951527051, 0.8760314339275924)
175
(-0.14525263773881875, 12.794578099583648, 0.7336448119254712)
325
(-0.3176123823473712, 14.165611588171958, 0.716007867152779)
350
(-0.5702550439633481, 15.906722481262415, 0.8484253024453171)
375
(-0.27617300985585835, 13.7757893415965, 0.7133847040932935)
450
(-0.842925853756705, 18.00608136379243, 0.6845390074594049)
475
(-0.5447896856752955, 16.111531521066865, 0.7185216899437274)
500
(-0.5276719332956944, 16.086657918718224, 0.9861786285315916)
525
(-0.4574264645955697, 15.550376231507796, 0.8365849491558203)
550
(-0.26787462799685535, 14.021859990238074, 0.7105219943028608)
625
(-0.05712317418077348, 11.669426779676051, 0.6081043705945822)
650
(0.19695304349504775, 8.263786632407742, 0.791781566459365)






    




---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-60-6ea98e28c74d> in <module>()
      1 for i in range(40):
      2     offset = i * 25
----> 3     regress = scipy.stats.linregress(absolute_humidity[offset:offset+50], np.log(air_value[offset:offset+50]))
      4     if (regress.rvalue**2 > 0.6):
      5         print(offset)

//anaconda/lib/python2.7/site-packages/scipy/stats/_stats_mstats_common.pyc in linregress(x, y)
     91 
     92     # average sum of squares:
---> 93     ssxm, ssxym, ssyxm, ssym = np.cov(x, y, bias=1).flat
     94     r_num = ssxym
     95     r_den = np.sqrt(ssxm * ssym)

/usr/local/lib/python2.7/site-packages/numpy/lib/function_base.pyc in cov(m, y, rowvar, bias, ddof, fweights, aweights)
   2259         if not rowvar and y.shape[0] != 1:
   2260             y = y.T
-> 2261         X = np.concatenate((X, y), axis=0)
   2262 
   2263     if ddof is None:

ValueError: all the input array dimensions except for the concatenation axis must match exactly

In [63]:

    

scipy.stats.linregress(absolute_humidity[:600], np.log(air_value[:600]))


    

    
Out[63]:





LinregressResult(slope=-0.21747856090248932, intercept=13.652655460958453, rvalue=-0.49729747618050696, pvalue=8.389654472566745e-39, stderr=0.01551526445474296)

In [105]:

    

plt.plot(np.log(air_value[:600]) - (humidity_value[:600] * -0.06568911456870866 + 17.20480030645947))
plt.plot(np.log(air_value[:600]) - (humidity_value[:600] * -0.04557760735953077 + 15.977729131340379))
plt.plot(np.log(air_value[:600]) - 13)


    

    
Out[105]:





[<matplotlib.lines.Line2D at 0x1c21e13650>]






    

In [104]:

    

plt.plot(np.log(air_value[:600]))
plt.plot(np.log(air_value[:600]) - (humidity_value[:600] * -0.06568911456870866 + 17.20480030645947))
plt.plot(np.log(air_value[:600]) - (humidity_value[:600] * -0.04557760735953077 + 15.977729131340379))


    

    
Out[104]:





[<matplotlib.lines.Line2D at 0x1c21ce9c50>]






    

In [40]:

    

import pandas as pd


    

In [109]:

    

#plt.plot(np.log(air_value[:600]) - (absolute_humidity[:600] * -0.5276719332956944 + 16.086657918718224))
#plt.plot(np.log(air_value[:600]) - (absolute_humidity[:600] * -0.21747856090248932 + 13.652655460958453))

plt.plot(np.log(air_value[:600]) - 13)


    

    
Out[109]:





[<matplotlib.lines.Line2D at 0x1c2239ea10>]






    

In [49]:

    

air_time[600]


    

    
Out[49]:





u'2019-01-21T19:15:37.93333Z'

In [50]:

    

humidity_time[600]


    

    
Out[50]:





u'2019-01-21T19:15:37.330998Z'

In [72]:

    

df = pd.DataFrame([i for i in r if i['measurement'] == 'max_volume'])


    

In [74]:

    

df['time'] = pd.to_datetime(df['timestamp'])


    

In [76]:

    

df.index = df['time']


    

In [99]:

    

plt.plot(df.groupby((df.index.hour + 17) % 24).median()['value'])


    

    
Out[99]:





[<matplotlib.lines.Line2D at 0x1a2062bf10>]






    

In [95]:

    

plt.plot(df['time'],scipy.signal.medfilt(df['value'], 17))


    

    
Out[95]:





[<matplotlib.lines.Line2D at 0x1a20804f50>]






    

In [119]:

    

plt.plot(depth[:100])


    

    
Out[119]:





[<matplotlib.lines.Line2D at 0x1a2073f890>]






    

In [ ]: