信号压缩矢量化V2,将时序信号转为B85字符串，100倍压缩率，python源码(原创算法，使用请引用)

 

原始数据：

[-1.13443950e-01 -1.45878432e-01 -1.79049990e-01 -2.12539910e-01  -2.45928442e-01 -2.78799573e-01 -3.10745720e-01 -3.41372281e-01  -3.70301996e-01 -3.97179073e-01 -4.21673013e-01 -4.43482116e-01  -4.62336608e-01 -4.78001364e-01 -4.90278207e-01 -4.99007732e-01  -5.04070674e-01 -5.05388780e-01 -5.02925189e-01 -4.96684324e-01  -4.86711297e-01 -4.73090844e-01 -4.55945795e-01 -4.35435125e-01  -4.11751590e-01 -3.85118995e-01 -3.55789131e-01 -3.24038412e-01  -2.90164261e-01 -2.54481303e-01 -2.17317385e-01 -1.79009507e-01  -1.39899688e-01 -1.00330827e-01 -6.06426160e-02 -2.11675442e-02   1.77729578e-02  5.58721816e-02  9.28415780e-02  1.28413923e-01   1.62346180e-01  1.94422024e-01  2.24454010e-01  2.52285342e-01   2.77791257e-01  3.00879980e-01  3.21493271e-01  3.39606533e-01   3.55228520e-01  3.68400620e-01  3.79195746e-01  3.87716854e-01   3.94095097e-01  3.98487660e-01  4.01075290e-01  4.02059569e-01   4.01659961e-01  4.00110663e-01  3.97657329e-01  3.94553671e-01   3.91058019e-01  3.87429850e-01  3.83926361e-01  3.80799096e-01   3.78290699e-01  3.76631809e-01  3.76038146e-01  3.76707824e-01   3.78818907e-01  3.82527254e-01  3.87964672e-01  3.95237379e-01   4.04219226e-01  4.13343882e-01  4.24570706e-01  4.37987680e-01   4.53655629e-01  4.71606864e-01  4.91844150e-01  5.14340045e-01   5.39036583e-01  5.65845351e-01  5.94647928e-01  6.25296720e-01   6.57616151e-01  6.91404228e-01  7.26434455e-01  7.62458071e-01   7.99206597e-01  8.36394668e-01  8.73723103e-01  9.10882193e-01   9.47555167e-01  9.83421789e-01  1.01816205e+00  1.05145989e+00   1.08300697e+00  1.11250637e+00  1.13967614e+00  1.16425286e+00   1.18599490e+00  1.20468549e+00  1.22013554e+00  1.23218620e+00   1.24071101e+00  1.24561775e+00  1.24684986e+00  1.24438750e+00   1.23824816e+00  1.22848680e+00  1.21519558e+00  1.19850317e+00   1.17857355e+00  1.15560445e+00  1.12982534e+00  1.10149497e+00   1.07089869e+00  1.03834527e+00  1.00416350e+00  9.68698523e-01   9.32307897e-01  8.95357548e-01  8.58217534e-01  8.21257756e-01   7.84843644e-01  7.49331856e-01  7.15066061e-01  6.82372851e-01   6.51557835e-01  6.22901955e-01  5.96658089e-01  5.73047966e-01   5.52259442e-01  5.34444181e-01  5.19715751e-01  5.08148186e-01   4.99775015e-01  4.94588789e-01  4.92541107e-01  4.93543144e-01   4.97466683e-01  5.04145641e-01  5.13378069e-01  5.24928616e-01   5.38531414e-01  5.53893371e-01  5.70697813e-01  5.88608458e-01   6.07273649e-01  6.26330814e-01  6.45411098e-01  6.64144099e-01   6.82162665e-01  6.99107684e-01  7.14632813e-01  7.28409081e-01   7.40129312e-01  7.49512315e-01  7.56306779e-01  7.60294830e-01   7.61295197e-01  7.59165950e-01  7.53806768e-01  7.45160709e-01   7.33215447e-01  7.18003975e-01  6.99604733e-01  6.78141180e-01   6.53780797e-01  6.26733525e-01  5.97249669e-01  5.65617266e-01   5.32158960e-01  4.97228422e-01  4.61206333e-01  4.24495997e-01   3.87518627e-01  3.50708354e-01  3.14507024e-01  2.79358841e-01   2.45704919e-01  2.13977816e-01  1.84596107e-01  1.57959068e-01   1.34441531e-01  1.14388987e-01  9.81129815e-02  8.58868749e-02   7.79420140e-02  7.44643694e-02  7.55916818e-02  8.14111553e-02   9.19577331e-02  1.07212978e-01  1.27104580e-01  1.51506499e-01   1.80239749e-01  2.13073818e-01  2.49728710e-01  2.89877600e-01   3.33150056e-01  3.79135814e-01  4.27389050e-01  4.77433117e-01   5.28765669e-01  5.80864151e-01  6.33191546e-01  6.85202361e-01   7.36348736e-01  7.86086643e-01  8.33882074e-01  8.79217158e-01   9.21596127e-01  9.60551067e-01  9.95647374e-01  1.02648885e+00   1.05272238e+00  1.07404215e+00  1.09019329e+00  1.10097495e+00   1.10624281e+00  1.10591087e+00  1.09995258e+00  1.08840127e+00   1.07134993e+00  1.04895019e+00  1.02141068e+00  9.88994652e-01   9.52017015e-01  9.10840667e-01  8.65872294e-01  8.17557617e-01   7.66376160e-01  7.12835592e-01  6.57465720e-01  6.00812198e-01   5.43430021e-01  4.85876887e-01  4.28706516e-01  3.72461982e-01   3.17669166e-01  2.64830392e-01  2.14418332e-01  1.66870247e-01   1.22582650e-01  8.19064436e-02  4.51426021e-02  1.25384474e-02  -1.57154267e-02 -3.94875454e-02 -5.87072248e-02 -7.33655017e-02  -8.35153048e-02 -8.92708620e-02 -9.08063466e-02 -8.83537756e-02  -8.22001808e-02 -7.26840830e-02 -6.01913103e-02 -4.51502053e-02  -2.80262787e-02 -9.31636987e-03  1.04576164e-02  3.07553231e-02   5.10248252e-02  7.07097027e-02  8.92562169e-02  1.06120503e-01   1.20775690e-01  1.32718866e-01  1.41477789e-01  1.46617282e-01   1.47745207e-01  1.44517962e-01  1.36645422e-01  1.23895258e-01   1.06096585e-01  8.31428812e-02  5.49941477e-02  2.16782642e-02  -1.67084681e-02 -6.00016169e-02 -1.07969742e-01 -1.60316346e-01  -2.16682629e-01 -2.76651026e-01 -3.39749490e-01 -4.05456464e-01  -4.73206505e-01 -5.42396474e-01 -6.12392246e-01 -6.82535842e-01  -7.52152911e-01 -8.20560477e-01 -8.87074847e-01 -9.51019604e-01  -1.01173357e+00 -1.06857867e+00 -1.12094756e+00 -1.16827098e+00  -1.21002471e+00 -1.24573604e+00 -1.27498965e+00 -1.29743297e+00  -1.31278067e+00 -1.32081856e+00 -1.32140661e+00 -1.31448107e+00  -1.30005580e+00 -1.27822268e+00 -1.24915106e+00 -1.21308638e+00  -1.17034782e+00 -1.12132514e+00 -1.06647463e+00 -1.00631430e+00  -9.41418331e-01 -8.72410787e-01 -7.99958803e-01 -7.24765181e-01  -6.47560564e-01 -5.69095254e-01 -4.90130770e-01 -4.11431247e-01  -3.33754776e-01 -2.57844790e-01 -1.84421600e-01 -1.14174178e-01  -4.77522851e-02  1.42409504e-02  7.12559285e-02  1.22802852e-01   1.68457307e-01  2.07865082e-01  2.40746158e-01  2.66897828e-01   2.86196906e-01  2.98600977e-01  3.04148702e-01  3.02959140e-01   2.95230116e-01  2.81235628e-01  2.61322348e-01  2.35905232e-01   2.05462300e-01  1.70528651e-01  1.31689772e-01  8.95742207e-02   4.48457790e-02 -1.80484198e-03 -4.96686404e-02 -9.80272036e-02  -1.46161723e-01 -1.93362035e-01 -2.38935577e-01 -2.82216150e-01  -3.22572388e-01 -3.59415811e-01 -3.92208389e-01 -4.20469493e-01  -4.43782176e-01 -4.61798684e-01 -4.74245125e-01 -4.80925260e-01  -4.81723335e-01 -4.76605936e-01 -4.65622832e-01 -4.48906792e-01  -4.26672374e-01 -3.99213699e-01 -3.66901217e-01 -3.30177528e-01  -2.89552272e-01 -2.45596168e-01 -1.98934261e-01 -1.50238453e-01  -1.00219414e-01 -4.96179540e-02  8.04027227e-04  5.02729168e-02   9.80129741e-02  1.43256011e-01  1.85251054e-01  2.23273862e-01   2.56636195e-01  2.84694718e-01  3.06859417e-01  3.22601446e-01   3.31460288e-01  3.33050148e-01  3.27065497e-01  3.13285691e-01   2.91578609e-01  2.61903261e-01  2.24311321e-01  1.78947569e-01   1.26049232e-01  6.59442095e-02 -9.51774445e-04 -7.41390931e-02  -1.53039152e-01 -2.37000291e-01 -3.25304590e-01 -4.17175501e-01  -5.11786210e-01 -6.08268658e-01 -7.05723084e-01 -8.03228012e-01  -8.99850536e-01 -9.94656811e-01 -1.08672260e+00 -1.17514378e+00  -1.25904665e+00 -1.33759792e+00 -1.41001436e+00 -1.47557177e+00  -1.53361345e+00 -1.58355780e+00 -1.62490515e+00 -1.65724364e+00  -1.68025406e+00 -1.69371373e+00 -1.69749912e+00 -1.69158748e+00  -1.67605717e+00 -1.65108691e+00 -1.61695378e+00 -1.57403012e+00  -1.52277927e+00 -1.46375032e+00 -1.39757168e+00 -1.32494392e+00  -1.24663160e+00 -1.16345437e+00 -1.07627746e+00 -9.86001572e-01  -8.93552372e-01 -7.99869649e-01 -7.05896329e-01 -6.12567429e-01  -5.20799095e-01 -4.31477871e-01 -3.45450298e-01 -2.63513005e-01  -1.86403375e-01 -1.14790927e-01 -4.92695084e-02  9.64960961e-03   6.15436418e-02  1.06083134e-01  1.43035679e-01  1.72268465e-01   1.93749607e-01  2.07548257e-01  2.13833481e-01  2.12871916e-01   2.05024227e-01  1.90740411e-01  1.70553992e-01  1.45075189e-01   1.14983118e-01  8.10171369e-02  4.39674180e-02  4.66488072e-03  -3.60294131e-02 -7.72352662e-02 -1.18064624e-01 -1.57632986e-01  -1.95070823e-01 -2.29534836e-01 -2.60218959e-01 -2.86364925e-01  -3.07272309e-01 -3.22307887e-01 -3.30914222e-01 -3.32617355e-01  -3.27033502e-01 -3.13874687e-01 -2.92953217e-01 -2.64184959e-01  -2.27591357e-01 -1.83300175e-01 -1.31544933e-01 -7.26630546e-02  -7.09273125e-03  6.46314656e-02  1.41884170e-01  2.23955950e-01   3.10061101e-01  3.99346417e-01  4.90900832e-01  5.83765837e-01   6.76946529e-01  7.69423182e-01  8.60163188e-01  9.48133244e-01   1.03231162e+00  1.11170039e+00  1.18533742e+00  1.25230809e+00   1.31175642e+00  1.36289569e+00  1.40501823e+00  1.43750433e+00   1.45983026e+00  1.47157511e+00  1.47242651e+00  1.46218509e+00   1.44076767e+00  1.40820908e+00  1.36466262e+00  1.31039912e+00   1.24580467e+00  1.17137695e+00  1.08772027e+00  9.95539322e-01   8.95631772e-01  7.88879717e-01  6.76240135e-01  5.58734433e-01   4.37437213e-01  3.13464375e-01  1.87960720e-01  6.20871691e-02  -6.29922312e-02 -1.86123369e-01 -3.06175046e-01 -4.22051710e-01  -5.32705796e-01 -6.37149537e-01 -7.34466081e-01 -8.23819804e-01  -9.04465682e-01 -9.75757617e-01 -1.03715560e+00 -1.08823166e+00  -1.12867444e+00 -1.15829251e+00 -1.17701615e+00 -1.18489778e+00  -1.18211091e+00 -1.16894766e+00 -1.14581490e+00 -1.11322898e+00  -1.07180919e+00 -1.02227000e+00 -9.65412131e-01 -9.02112624e-01  -8.33313985e-01 -7.60012539e-01 -6.83246146e-01 -6.04081409e-01  -5.23600539e-01 -4.42888027e-01 -3.63017283e-01 -2.85037412e-01  -2.09960272e-01 -1.38747985e-01 -7.23010423e-02 -1.14471539e-02   4.30690206e-02  9.05951381e-02  1.30579316e-01  1.62577054e-01   1.86256853e-01  2.01404444e-01  2.07925595e-01  2.05847432e-01   1.95318281e-01  1.76606003e-01  1.50094858e-01  1.16280917e-01   7.57660840e-02  2.92507909e-02 -2.24745546e-02 -7.85392586e-02  -1.36150080e-01 -1.95662574e-01 -2.56493366e-01 -3.17567385e-01  -3.77790504e-01 -4.36063310e-01 -4.91295016e-01 -5.42417336e-01  -5.88398164e-01 -6.28254886e-01 -6.61067162e-01 -6.85989028e-01  -7.02260162e-01 -7.09216176e-01 -7.06297810e-01 -6.93058907e-01  -6.69173070e-01 -6.34438919e-01 -5.88783869e-01 -5.32266392e-01  -4.65076705e-01 -3.87535903e-01 -3.00093506e-01 -2.03323464e-01  -9.79186525e-02  1.53160754e-02  1.35472218e-01  2.61547224e-01   3.92455266e-01  5.27039063e-01  6.64082602e-01  8.02324638e-01   9.40472787e-01  1.07721807e+00  1.21124972e+00  1.34127007e+00   1.46600944e+00  1.58424060e+00  1.69479303e+00  1.79656636e+00   1.88854320e+00  1.96980101e+00  2.03952290e+00  2.09700723e+00   2.14167595e+00  2.17308151e+00  2.19091223e+00  2.19499622e+00   2.18530356e+00  2.16194690e+00  2.12518046e+00  2.07539724e+00   2.01312478e+00  1.93901921e+00  1.85385788e+00  1.75853054e+00   1.65402922e+00  1.54143685e+00  1.42191491e+00  1.29669003e+00   1.16703997e+00  1.03427883e+00  8.99741997e-01  7.64770686e-01   6.30696537e-01  4.98826252e-01  3.70426558e-01  2.46709640e-01   1.28819225e-01  1.78174813e-02 -8.53271007e-02 -1.79750717e-01  -2.64704000e-01 -3.39560395e-01 -4.03823070e-01 -4.57130259e-01  -4.99258991e-01 -5.30127137e-01 -5.49793770e-01 -5.58457812e-01  -5.56454994e-01 -5.44253164e-01 -5.22445981e-01 -4.91745092e-01  -4.52970864e-01 -4.07041793e-01 -3.54962709e-01 -2.97811918e-01  -2.36727437e-01 -1.72892485e-01 -1.07520406e-01 -4.18392007e-02   2.29241429e-02  8.55593312e-02  1.44888269e-01  1.99780201e-01   2.49166299e-01  2.92053524e-01  3.27537591e-01  3.54814876e-01   3.73193115e-01  3.82100770e-01  3.81094936e-01  3.69867692e-01   3.48250819e-01  3.16218811e-01  2.73890152e-01  2.21526820e-01   1.59532035e-01  8.84462613e-02  8.94150858e-03 -7.81860030e-02  -1.72024735e-01 -2.71558179e-01 -3.75676531e-01 -4.83189580e-01  -5.92840634e-01 -7.03321330e-01 -8.13287170e-01 -9.21373566e-01  -1.02621223e+00 -1.12644771e+00 -1.22075387e+00 -1.30785009e+00  -1.38651705e+00 -1.45561188e+00 -1.51408244e+00 -1.56098066e+00  -1.59547467e+00 -1.61685966e+00 -1.62456726e+00 -1.61817343e+00  -1.59740456e+00 -1.56214199e+00 -1.51242460e+00 -1.44844962e+00  -1.37057160e+00 -1.27929948e+00 -1.17529191e+00 -1.05935073e+00  -9.32412795e-01 -7.95540176e-01 -6.49908873e-01 -4.96796143e-01]

 

压缩后数据：

IO+w!r7w~^Zf8?nfR2cbs==_3kEEM;ccs0SYIBBPBrbK5XHR>SQwSn>id$y9)_5UAjdnY0=l_mDbDmBBGFb

import numpy as np
from matplotlib import pyplot as plt
import base64

def init_N(n, k, X, T):
    """
    生成初始化的参数矩阵
    :param n: 控制点数量，int
    :param k: 基函数的阶数，int, 默认为2，建议2,3,4,最小为2，越大越平滑，越大速度越慢
    :param X: 自变量矩阵，取值范围[0,1] (shape=[sample], dtype=float)
    :param T: 控制点系数矩阵，取值范围[0,1] (shape=[n + k + 1], dtype=float)
    :return N:  (shape=[sample, n], dtype=float)
    """ 
    m = n + k + 1
    X = X[None]    # shape=[1, sample] 
    T = T[:, None] # shape=[n+k+1, 1] 
    m = n + k + 1
    N = 1.0 * (T[:-1] <= X) * (X <= T[1:])  # shape=[n+k, sample], dtype=float
    for ki in range(1, k + 1):  # 动态规划 k ∈ [1, k]
        n = m - ki - 1  # 每次更新的控制点数量
        c1 = T[ki:n + ki] - T[:n]
        c2 = T[ki + 1:n + ki + 1] - T[1:n + 1,]
        c1[c1 == 0] = 1 # 替换 c1 等于 0 的情况，如果等于0 则替换成1，否则不变
        c2[c2 == 0] = 1 # 替换 c2 等于 0 的情况，如果等于0 则替换成1，否则不变
        w1 = (X - T[:n]) / c1
        w2 = (T[ki + 1:n + ki + 1] - X) / c2
        N = N[:-1] * w1 + N[1:] * w2
    N = N[:n].T # 截取前n个，然后交换行列
    return N  # sample, n


def init_T_X(k, m, sample):
    """
    生成初始化的参数向量
    :param k: 基函数的阶数，int, 默认为2，建议2,3,4,最小为2，越大越平滑，越大速度越慢
    :param m: 系数矩阵的长度，int，m=n+k+1
    :param sample: 采样点数量， int
    :return T: 系数矩阵，取值范围[0,1] (shape[n + k + 1], dtype=float)
    :return X: 自变量矩阵，取值范围[0,1] (shape=[sample], dtype=float)
    """
    # clame 模式
    T = np.concatenate([
        np.zeros(k),
        np.linspace(0, 1, m - 2 * k),
        np.ones(k)]
    )
    gama = 1e-9
    first = T[k]
    last = T[-k - 1] * (1 - gama) + gama * T[-k - 2]
    X = np.linspace(first, last, sample)
    return T, X


def unzip(P, k, sample):
    """
    解压点，生成曲线
    :param P: 控制点矩阵，(shape=[n], dtype=float)
    :param k: 基函数的阶数，int, 默认为2，建议2,3,4,最小为2，越大越平滑，越大速度越慢
    :param sample: 采样点数量， int
    :return data: 解压后的曲线，(shape=[sample], dtype=float)
    """
    n = len(P)
    T, X = init_T_X(k=k, m=k + n + 1, sample=sample)
    N = init_N(n=n, k=k, T=T, X=X) 
    numerator = np.sum(P[None] * N, axis=-1)
    denominator = np.sum(N, axis=-1)
    data = numerator / denominator
    return data
 

def zip(data, n, k):
    """
    压缩曲线，生成控制点矩阵
    :param data: 曲线上的采样点，(shape=[sample], dtype=float)
    :param n: 控制点的数量，int
    :param k: 曲线的阶数，int
    :return P: 控制点矩阵， （shape=[n], dtype=float）
    """
    sample = len(data)
    T, X = init_T_X(k=k, m=k + n + 1, sample=sample) 
    a = init_N(n=n, k=k, T=T, X=X)  
    b = data * np.sum(a, axis=-1) 
    P = np.linalg.lstsq(a, b, rcond=None)[0]
    loss = np.sqrt(np.sum((a@P[:,None] - b[:, None])**2)) # 求二范数
    print('loss : ', loss)
    return P,loss

def plot_n_loss(data, k=3):
    ns,losss = [],[]
    for i in range(10, 128):
        P, loss = zip(data, i, k)
        ns.append(i)
        losss.append(np.log10(loss))
    plt.plot(ns, losss)
    plt.show()

def zip_auto(data,
             min_n = 4,
             max_n = 128,
             loss_threshold= 0.1,# 可以接收的最小误差范围
             ):
    """
    自适应压缩曲线，生成控制点矩阵，自动查找合适的n
    :param data: 曲线上的采样点，(shape=[sample], dtype=float)
    :param min_n: 最小的控制点数量，int
    :param max_n: 最大的控制点数量，int 
    :param min_loss_threshold: 最小的二范数误差，float
    :param max_loss_threshold: 最大的二范数误差，float
    :return P: 控制点矩阵， （shape=[n], dtype=float）
    :return loss: 二范数误差， （shape=[n], dtype=float）
    """
    k = 3   
    l,r = min_n - 1, max_n + 1
    n = l
    resP = None
    resLoss = None
    while l < r*0.9: # r越大允许的误差越大，r越小允许的n的误差范围越小
        n = (l+r)//2
        P, loss = zip(data, n, k)  
        if loss > loss_threshold:
            l = n + 1
        else:
            resP = P
            resLoss = loss
            r = n - 1
            if not l < r*0.85:
                break
    if resP is None:
        resP = P
        resLoss = loss 
    return resP, resLoss
 
def unzip_auto(P, sample):
    """
    自动解压点，生成曲线
    :param P: 控制点矩阵，(shape=[n], dtype=float)
    :param sample: 采样点数量， int
    :return data: 解压后的曲线，(shape=[sample], dtype=float)
    """
    k = 3
    n = len(P)
    T, X = init_T_X(k=k, m=k + n + 1, sample=sample)
    N = init_N(n=n, k=k, T=T, X=X) 
    numerator = np.sum(P[None] * N, axis=-1)
    denominator = np.sum(N, axis=-1)
    data = numerator / denominator
    return data
 

class LineBytes:
    def __init__(self, half=False):
        """ 初始化构造函数
        :param half: 是否使用半精度浮点数，bool
        """
        if half:
            self.dtype = np.float16
            self.head_bytes = 4
        else:
            self.dtype = np.float32
            self.head_bytes = 8

    def unzip(self, bytes, sample): 
        head = np.frombuffer(bytes[:self.head_bytes], dtype=self.dtype)
        map = np.frombuffer(bytes[self.head_bytes:], dtype=np.uint8).astype(np.float32)
        points = head[0] + map * head[1]  # 后面的数据做映射解码 
        data = unzip_auto(points, sample)
        return data, points

    def zip(self, data):
        """ 压缩函数
        :param data: 需要压缩的数据列表，(shape=[-1], dtype=float)
        :return b85: 压缩后的base85编码的字符串，str
        """
        points, loss = zip_auto(data, loss_threshold=0.1) 
        head = np.array([points.min(), (points.max() - points.min()) / 255.])
        map = (points - head[0]) / head[1]
        bytes = head.astype(self.dtype).tobytes() + map.astype(np.uint8).tobytes() 
        return bytes
    
    def unzipb85(self, b85:str, sample=100):
        """ 解压缩函数
        :param b85: base85编码的字符串，str
        :return data: 解压缩后的数据 (shape=[-1], dtype=float)
        """
        bytes = base64.b85decode(b85)
        return self.unzip(bytes, sample)

    def zipb85(self, data):
        """ 压缩函数
        :param data: 需要压缩的数据列表，(shape=[-1], dtype=float)
        :return b85: 压缩后的base85编码的字符串，str
        """ 
        b85 = base64.b85encode(self.zip(data))
        return b85
 

if __name__ == "__main__":
    sample = 720 # 采样数量
    X = np.linspace(0, 1, sample)  
    Y = np.sin(3 * np.pi * X + 6) * (0.2 + np.abs(X - 0.8)) + \
        np.sin(25 * np.pi * X + 9) * (0.2 + np.abs(X - 0.1)) + \
        np.sin(11 * np.pi * X + 3) * (0.2 + np.abs(X - 0.1)) 
    
    lb = LineBytes()
    b85 = lb.zipb85(Y)
    unzip_Y, points = lb.unzipb85(b85, sample) 
    print(Y)
    print(len(b85), b85)
    chars = b85.decode()
    print(len(chars), chars)
    print("mean/μ :", Y.mean())
    print("var/σ :", Y.var()) 
    print("err/ι :", np.abs(unzip_Y - Y).mean())
    
    plt.plot(X, Y, 'g-', marker='+', label='origin line') 
    plt.plot(np.linspace(0,1, len(points)), points, 'b-.', marker='o', label='zip points') 
    plt.plot(X, unzip_Y, 'r-', marker='+', label='unzip line') 
    plt.legend()
    plt.show()