Why does it become all zeros after flattening in NATR, while this doesn't happen in other APIs?

[gucasbrg]

import numpy as np
import talib

def _ta_natr_14(x1, x2, x3):
    t = 14
    x1 = x1.flatten()
    x2 = x2.flatten()
    x3 = x3.flatten()
    print(np.nan_to_num(talib.NATR(x1, x2, x3, timeperiod=t))[:100])

def _ta_natr_14_qs(x1, x2, x3):
    t = 14
    print(np.nan_to_num(talib.NATR(x1, x2, x3, timeperiod=t))[:100])


x1 = np.load('x1.npy')
x2 = np.load('x2.npy')
x3 = np.load('x3.npy')
_ta_natr_14(x1, x2, x3)
_ta_natr_14_qs(x1, x2, x3)

data.zip

[mrjbq7]

It doesn't become all zeros, but I'm not sure why the initial values are ~0, have to look at it closer maybe.

In [1]: import numpy as np

In [2]: import talib as ta

In [3]: x1 = np.load('x1.npy')
   ...: x2 = np.load('x2.npy')
   ...: x3 = np.load('x3.npy')
   ...: 

In [4]: r = ta.NATR(x1, x2, x3, timeperiod=14)

In [6]: r
Out[6]: 
array([          nan,           nan,           nan, ..., -252.67133272,
       -241.77978634, -222.86091276])

In [7]: r[:100]
Out[7]: 
array([            nan,             nan,             nan,             nan,
                   nan,             nan,             nan,             nan,
                   nan,             nan,             nan,             nan,
                   nan,             nan, 0.00000000e+000, 2.15081535e-314,
       2.15084699e-314, 2.15081535e-314, 2.38508723e-314, 2.38509025e-314,
       2.15358054e-314, 2.15084699e-314, 2.15358054e-314, 2.38508794e-314,
       2.38508621e-314, 2.15085935e-314, 2.38508462e-314, 2.15085935e-314,
       2.38508804e-314, 2.15357763e-314, 2.15084699e-314, 2.38508555e-314,
       2.38508851e-314, 2.15082759e-314, 2.15084699e-314, 2.15082759e-314,
       2.38508968e-314, 2.38508612e-314, 2.14447468e-314, 2.15084699e-314,
       2.14447468e-314, 2.38508859e-314, 2.38508934e-314, 2.15075115e-314,
       2.15084699e-314, 2.15075115e-314, 2.38508427e-314, 2.38508714e-314,
       2.15077545e-314, 2.15084699e-314, 2.15077545e-314, 2.38508560e-314,
       2.38508984e-314, 2.15358013e-314, 2.15084699e-314, 2.15358013e-314,
       2.38508956e-314, 2.38508536e-314, 2.14564750e-314, 2.15084699e-314,
       2.14564750e-314, 2.38508740e-314, 2.38508882e-314, 2.15079954e-314,
       2.15084699e-314, 2.15079954e-314, 2.38508702e-314, 2.38508602e-314,
       2.15059941e-314, 2.15084699e-314, 2.15059941e-314, 2.38508657e-314,
       2.38508885e-314, 2.14868621e-314, 2.15084699e-314, 2.14868621e-314,
       2.38508531e-314, 2.38508674e-314, 2.15080767e-314, 2.15084699e-314,
       2.15080767e-314, 2.38508918e-314, 2.38508638e-314, 2.15082959e-314,
       2.15084699e-314, 2.15082959e-314, 2.38508799e-314, 2.38508662e-314,
       2.15086285e-314, 2.15084699e-314, 2.15086285e-314, 2.38508517e-314,
       2.38508735e-314, 2.14662694e-314, 2.38508631e-314, 2.14662694e-314,
       2.38508690e-314, 2.14762244e-314, 2.15084699e-314, 2.38508823e-314])

In [8]: r[100:200]
Out[8]: 
array([ 2.38508633e-314,  2.15072236e-314,  2.38508538e-314,
        2.15072236e-314,  2.38508749e-314,  2.15086156e-314,
        2.38509105e-314,  2.38509108e-314,  2.15483403e-314,
        2.15084699e-314,  2.38509110e-314,  2.38509112e-314,
        2.14564861e-314,  2.38509115e-314,  2.14564861e-314,
        2.38509117e-314,  2.15078548e-314,  2.15084699e-314,
        2.38509119e-314,  2.38509122e-314,  2.14564864e-314,
        2.38509124e-314,  2.14564864e-314,  2.38509127e-314,
        2.14564583e-314,  2.15084699e-314,  2.38509129e-314,
        2.38509131e-314,  2.15073361e-314,  2.38509134e-314,
        2.15073361e-314,  2.38509136e-314,  2.15072353e-314,
        2.15084699e-314,  2.38509138e-314,  2.38509141e-314,
        2.15073361e-314,  2.38509143e-314,  2.15073361e-314,
        2.38509145e-314,  2.14564618e-314,  2.38509148e-314,
        2.38509150e-314,  2.15491235e-314,  2.15084805e-314,
        2.38509153e-314,  2.14712372e-314,  2.38509155e-314,
        2.38509157e-314,  2.15491251e-314,  2.15084805e-314,
        2.38509160e-314,  2.14712293e-314,  2.38509162e-314,
        2.38509164e-314,  2.15491238e-314,  2.15084699e-314,
        2.38509167e-314,  2.38509169e-314,  2.15073361e-314,
        2.38509172e-314,  2.15073361e-314,  2.38509174e-314,
        7.27720355e+001,  7.11351962e+001,  6.96721234e+001,
        7.04181050e+001,  7.54645854e+001,  8.55322620e+001,
        9.53305916e+001,  1.04393368e+002,  1.10398671e+002,
        1.11875702e+002,  1.13083248e+002,  1.14092627e+002,
        1.15133525e+002,  1.16161142e+002,  1.23706637e+002,
        1.31069486e+002,  1.38575068e+002,  1.46799978e+002,
        1.49581612e+002,  1.55800334e+002,  1.69201162e+002,
        2.07889468e+002,  7.58211146e+002,  1.12955155e+004,
        1.65019307e+002,  1.32951872e+002,  1.24512123e+002,
        1.32892189e+002,  1.47207693e+002,  1.69148266e+002,
        2.04277895e+002,  2.67542693e+002,  4.09486356e+002,
        7.09521998e+002,  9.90584944e+002,  6.69927309e+002,
       -3.94789948e+003])

[mrjbq7]

In [7]: r = ta.NATR(x1.flatten(), x2.flatten(), x3.flatten(), timeperiod=14)

In [8]: r[:100]
Out[8]: 
array([nan, nan, nan, nan, nan, nan, nan, nan, nan, nan, nan, nan, nan,
       nan,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,
        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,
        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,
        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,
        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,
        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,
        0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.])

In [9]: r[100:200]
Out[9]: 
array([    0.        ,     0.        ,     0.        ,     0.        ,
           0.        ,     0.        ,     0.        ,     0.        ,
           0.        ,     0.        ,     0.        ,     0.        ,
           0.        ,     0.        ,     0.        ,     0.        ,
           0.        ,     0.        ,     0.        ,     0.        ,
           0.        ,     0.        ,     0.        ,     0.        ,
           0.        ,     0.        ,     0.        ,     0.        ,
           0.        ,     0.        ,     0.        ,     0.        ,
           0.        ,     0.        ,     0.        ,     0.        ,
           0.        ,     0.        ,     0.        ,     0.        ,
           0.        ,     0.        ,     0.        ,     0.        ,
           0.        ,     0.        ,     0.        ,     0.        ,
           0.        ,     0.        ,     0.        ,     0.        ,
           0.        ,     0.        ,     0.        ,     0.        ,
           0.        ,     0.        ,     0.        ,     0.        ,
           0.        ,     0.        ,     0.        ,    72.77203545,
          71.13519618,    69.67212345,    70.41810497,    75.46458542,
          85.53226199,    95.33059159,   104.3933683 ,   110.39867053,
         111.87570231,   113.08324841,   114.09262749,   115.13352545,
         116.16114151,   123.70663687,   131.06948569,   138.57506813,
         146.79997804,   149.5816119 ,   155.80033379,   169.20116231,
         207.88946832,   758.21114577, 11295.51548872,   165.01930733,
         132.9518719 ,   124.51212282,   132.89218869,   147.20769277,
         169.14826558,   204.27789478,   267.54269313,   409.48635643,
         709.52199833,   990.58494426,   669.92730938, -3947.89947949])