In [1]:

    

%matplotlib inline 
import numpy as np
import pandas as pd
import re 
import itertools
import ast
import matplotlib.pyplot as plt


from sklearn.linear_model import  LinearRegression

positions = np.array([[6, 6],
                  [6, 18],
                  [6, 30],
                  [6, 42],
                  [18, 6],
                  [18, 18],
                  [18, 30],
                  [18, 42],
                  [30, 6],
                  [30, 18],
                  [30, 30],
                  [30, 42],
                  [42, 6],
                  [42, 18],
                  [42, 30],
                  [42, 42],
                  [54, 6],
                  [54, 18],
                  [54, 30],
                  [54, 42]])
numsensors = 6
numtrials = len(positions)

train_file = "../src/arduino/data/recolect_points_20160116_scrapping.txt"
test_file = "../src/arduino/data/mid_points_test_20160116_scrapping.txt"

pattern = "hit: {(.*)}"
hit_times = []
hit_values=[]
train_times = np.zeros([numtrials,numsensors])
train_values = np.zeros([numtrials,numsensors])
diff_pairs = numsensors*(numsensors-1)/2

##### TRAIN DATA AQUISITION
i = 0
with open(train_file) as fo:
    for line in fo:
        if line.startswith("("): # a new point is stored
            if len(hit_times)> 0 :
                array_time = np.array(hit_times)
                array_value = np.array(hit_values)
                plt.title(positions[i])
                plt.errorbar(range(1,numsensors+1),array_time.mean(axis=0), array_time.std(axis=0), linestyle='None', marker='^')
                plt.xlim([0.5, len(values) + 0.5])
                plt.show()
                train_times[i,:]= array_time.mean(axis=0)  #mean of hits in the same point
                train_values[i,:]= array_value.mean(axis=0)
                hit_times = [] #init hit_times list
                hit_values = [] #init hit_values
                i = i +1
        else:
            hit_re = re.match(pattern, line)
            if hit_re == None:
                continue
            sensors=hit_re.group(1).split(" ")
            times = []
            values = []
            for s in sensors[0:-1]:
                time,bd = s.split(":")
                times.append(int(time))
                values.append(int(bd))
            side = sensors[-1]
            hit_values.append(values)
            hit_times.append(times)


#ultim cas:  
mean_time = np.array(hit_times).mean(axis=0)
mean_value = np.array(hit_values).mean(axis=0)
train_times[i,:]= mean_time
train_values[i,:]= mean_value


# Matriu de diff per a train
train_diff= np.zeros([numtrials,diff_pairs])
for j in range(len(train_times)):
    i = 0
    for a in itertools.combinations(range(numsensors), 2):
        train_diff[j,i] = train_times[j,a[0]]-train_times[j,a[1]]
        i=i+1
        
train_df =  pd.DataFrame(train_times,index = range(1,numtrials+1) ,columns=range(1,numsensors+1))
train_y = pd.DataFrame(positions,columns=["X","Y"],index = range(1,numtrials+1))
train_df.join(train_y, how='inner')

diff_df = pd.DataFrame(train_diff,index = range(1,numtrials+1) ,columns=range(1,diff_pairs+1))
diff_df.join(train_y, how='inner')

for i,train in enumerate(train_values):
    plt.plot(train_times[i],label=positions[i])
    if (i % 4) == 3 :
        plt.title("Train")
        plt.legend(loc='upper left', bbox_to_anchor=(1, 1))
        plt.show()

    

##### Machine Learning Fit del train

l = LinearRegression()
l.fit(train_diff, positions)
dists_train = []
for i,train in enumerate(train_diff):
    dists_train.append(np.linalg.norm(l.predict(train)[0]-positions[i]))
   

print "Train Error"
print pd.DataFrame(dists_train, index = range(1,numtrials+1)).join(train_y, how='inner')


## Testing
dists_test = []
positions_test= []
predict_test=[]
current_position=[]
with open(test_file) as fo:
    for line in fo:    # For each line in test
        if line.startswith("["): # a new point is stored
            if (len(current_position)>0):
                plt.title(current_position)
                plt.legend(loc='upper left', bbox_to_anchor=(1, 1))
                plt.show()
   
            current_position = ast.literal_eval(line[0:-2])
        else:  
            hit_re = re.match(pattern, line)
            if hit_re == None:
                continue
            sensors=hit_re.group(1).split(" ")
            test_times = []
            test_values = []
            for s in sensors[0:-1]:
                time,bd = s.split(":")
                test_times.append(int(time))
                test_values.append(int(bd))
            
            side = sensors[-1]
            test= np.zeros(diff_pairs)
            for i,a in enumerate(itertools.combinations(range(numsensors), 2)):
                test[i] = test_times[a[0]]-test_times[a[1]]
            predict = l.predict(test)[0]
            predict_test.append(predict)    
            dists_test.append(np.linalg.norm(predict-current_position))
            positions_test.append(current_position)
            plt.plot(test_times,label=dists_test[-1])
           
#last case
plt.title(current_position)
plt.legend(loc='upper left', bbox_to_anchor=(1, 1))
plt.show()
        
print "Test Error"
print pd.DataFrame(predict_test,columns=["X_","Y_"]).join(pd.DataFrame(positions_test,columns=["X","Y"])).join(pd.DataFrame(dists_test,columns=["dists"]), how='inner')


    

    




/Users/eloi/anaconda2/lib/python2.7/site-packages/matplotlib/collections.py:590: FutureWarning: elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison
  if self._edgecolors == str('face'):






    













    













    













    













    













    













    













    













    













    













    













    













    













    













    













    













    













    













    













    













    













    













    













    













    




Train Error
            0   X   Y
1    3.084996   6   6
2    4.360477   6  18
3    2.503543   6  30
4    3.998490   6  42
5    8.079524  18   6
6    5.898015  18  18
7    4.644635  18  30
8   14.299352  18  42
9    0.868339  30   6
10   4.028435  30  18
11   0.797446  30  30
12  12.032358  30  42
13   4.164541  42   6
14   1.793049  42  18
15   0.995736  42  30
16   4.038379  42  42
17   8.166643  54   6
18   6.792805  54  18
19   9.392817  54  30
20   4.545266  54  42






    













    













    













    













    













    













    













    













    













    













    




/Users/eloi/anaconda2/lib/python2.7/site-packages/matplotlib/axes/_axes.py:475: UserWarning: No labelled objects found. Use label='...' kwarg on individual plots.
  warnings.warn("No labelled objects found. "






    













    













    




Test Error
           X_         Y_   X   Y      dists
0   14.308553  16.293247  12  12   4.874565
1   18.401630  20.670906  12  12  10.778009
2   16.302757  21.308799  12  12  10.255119
3   18.176911  20.036302  12  12  10.135896
4   14.754138  16.696002  12  12   5.444053
5   15.509394  21.806512  12  12  10.415543
6   15.253591  16.717142  12  12   5.730382
7   16.749648  15.611951  12  12   5.967021
8    6.070239  16.804960  12  24   9.323661
9    9.061325  18.984969  12  24   5.812602
10  10.657912  17.689283  12  24   6.451849
11  13.415846  18.643395  12  24   5.540562
12  10.533518  18.329067  12  24   5.857479
13   7.254158  15.245057  12  24   9.958516
14 -58.072918 -27.151223  12  36  94.330752
15   4.565683  34.498870  12  36   7.584356
16   4.132841  34.362164  12  36   8.035838
17  28.673597  17.827048  24  12   7.469739
18  27.711106  16.263875  24  12   5.652693
19  29.286012  18.938870  24  12   8.722949
20  28.573684  18.663960  24  12   8.082509
21  29.061878  16.632909  24  12   6.861957
22  30.963574  33.206535  24  24  11.543468
23  31.707040  33.524462  24  24  12.252096
24  30.542921  30.148046  24  24   8.978212
25  31.580970  33.390163  24  24  12.068400
26  30.357511  31.969723  24  24  10.194824
27  24.400505  40.623974  24  36   4.641287
28  19.858768  45.137874  24  36  10.032475
29  17.827395  42.289264  24  36   8.812259
30  12.416406  42.899492  24  36  13.482679
31  -9.008279  48.186400  24  36  35.186003
32  38.992256  14.526379  36  12   3.916145
33  39.301664  14.992458  36  12   4.455983
34  39.816281  14.762495  36  12   4.711197
35  37.680258  12.726947  36  12   1.830770
36  39.799657  12.815144  36  12   3.886110
37  39.593934  14.282916  36  12   4.257707
38  24.587220  49.307666  36  24  27.762016
39  30.593964  47.586638  36  24  24.198238
40  21.452502  39.433344  36  24  21.208909
41  29.669795  37.334131  36  24  14.760439
42  29.209573  36.787832  36  24  14.478900
43  37.934760  33.470467  36  24   9.666077
44  47.608681   9.519082  36  36  28.913673
45  48.302412  14.379541  36  36  24.875562
46  47.084063   9.148992  36  36  29.048805
47  47.246959   6.924696  36  36  31.174788
48  48.883563  12.819166  36  36  26.520506
49  45.273036  21.122000  48  12   9.520883
50  45.172722  20.807787  48  12   9.250438
51  44.648211  20.896652  48  12   9.507098
52  44.260202  21.290132  48  12  10.014621
53  48.130472  20.285250  48  12   8.286277
54  46.242247  28.002712  48  36   8.188181
55  44.286912  29.322258  48  36   7.640632
56  48.351893  28.453325  48  36   7.554874
57  44.035793  29.549279  48  36   7.571443
58  50.975291  34.979545  48  36   3.145423

In [ ]: