時系列データのマハラノビス距離行列を返したい。

Question

下のような50*5の時系列データ（rtn_insample）について、各列間のマハラノビス距離を行列で求めたいのですが、たとえばユークリッド距離行列なら簡単に、

from scipy.spatial import distance
dist=distance.cdist(rtn_insample.T, rtn_insample.T, "euclidean")

で求まりますが、マハラノビス距離行列を返すことがうまくできません。
（例えば、
cov=rtn_insample.T.cov()
dist=distance.cdist(rtn_insample.T, rtn_insample.T, "mahalanobis", VI=cov_inv)
とすると5*5の行列を返しますが、nanが多く時系列データを増やすとほとんどnanになります。）

何か簡単にできる方法ないでしょうか？

Aica Kogyo Company, Limited	Aisin Seiki Co Ltd	Ajinomoto Co., Inc.	ANA Holdings Inc.	Aiful Corporation

2016/10/21 0.001084208 -0.02004218 -0.00805223 0.013038754 0.003236294
2016/10/24 0.01768959 0.00107646 0.007240057 0.003575206 0.003225803
2016/10/25 0.005675793 0.01720428 0.007623553 0.008193851 0.009646297
2016/10/26 -0.005290985 -0.010570824 0.004107237 0.0017668 0.006369472
2016/10/27 -0.007801414 -0.003205121 0.009257317 -0.003174603 0.00316453
2016/10/28 0.08470333 -0.008574486 -0.000213325 0.02512383 0.006309152
2016/10/31 -0.011532128 -0.002162158 -0.002986968 0.01898515 0.003134847
2016/11/1 0.01833332 0.001083374 -0.001497984 0.012872577 0.003124952
2016/11/2 -0.0163666 -0.01731604 0.006643772 -0.01739132 -0.012461066
2016/11/3 0 0 0 0 0
2016/11/4 0 -0.00881058 -0.003193557 -0.008849561 -0.01577288
2016/11/7 0.008319497 0.01444447 0.002990127 0.007211566 0.01602566
2016/11/8 -0.003300309 0 -0.03982115 0.003750443 -0.006309152
2016/11/9 -0.03907287 -0.04162103 -0.08294524 -0.03091031 -0.05079365
2016/11/10 0.04583049 0.03771424 0.02370012 0.02769017 0.10033441
2016/11/11 -0.05996704 -0.00881058 -0.03000236 0.005457044 0.02127659
2016/11/14 0.03680336 0.02333331 -0.003409624 0.006105781 0.02976191
2016/11/15 -0.008451641 0.003257275 -0.005865097 -0.009777486 -0.0346821
2016/11/16 -0.012615085 -0.001082242 -0.001966596 0.015321732 0.04491019
2016/11/17 0.003453016 -0.013001084 0.0300492 0.01609659 0.02865326
2016/11/18 0.01238811 0.01646543 0.014108062 0.011551142 0.011142015
2016/11/21 0.005438447 0.007559419 0.03112471 0.014681935 0.01652896
2016/11/22 0.002366424 0.009646297 -0.002058089 0.000321508 0.010840058
2016/11/23 0 0 0 0 0
2016/11/24 0.001349092 0.014861941 -0.009165883 0.014786243 -0.010723889
2016/11/25 0.002020836 0.010460258 0.02358925 -0.002217293 -0.02439022
2016/11/28 0.010084033 0.00310564 -0.009037495 0.00444448 -0.002777755
2016/11/29 -0.0063228 0.008255959 0.003875971 0.003160596 0.002785563
2016/11/30 0.008037567 0.001023531 0.004996538 -0.00031507 -0.01944447
2016/12/1 -0.001661122 0.00920248 0.02214694 -0.02458239 -0.00566572
2016/12/2 -0.007653892 -0.012158036 -0.003979683 0.001615524 0.005697966
2016/12/5 -0.008718967 -0.02051282 -0.007991135 -0.012903214 -0.00283289
2016/12/6 -0.002368093 0.006282687 -0.02327144 0.000653625 0.002840877
2016/12/7 0.006103754 0.002081156 0.003436446 0.004572153 0.005665779
2016/12/8 0.02460396 0.0238837 0.01598179 0.01983094 0.1042254
2016/12/9 0.009868383 0.005071044 0.002921343 0.003187776 -0.03061223
2016/12/12 0.008143306 0.009081721 0.04638135 0.006355286 -0.013157904
2016/12/13 0.0080775 0.001999974 0.0126338 0.005367875 -0.02133334
2016/12/14 -0.02083331 0.005988 0.001480222 0.001570344 0
2016/12/15 -0.006546617 0.02182543 0.002533793 -0.005644381 -0.02179837
2016/12/16 0.01647449 0.015533924 0.010320187 0.01829076 -0.005571008
2016/12/19 0 -0.003824115 0.013133168 -0.000309706 0.00280118
2016/12/20 0.008103728 0.001919389 -0.009670794 0.003717422 -0.002793312
2016/12/21 -0.004823148 0.007662892 -0.008518577 -0.002469122 -0.011204481
2016/12/22 -0.008077562 0.0114069 -0.010477781 -0.002784669 -0.008498609
2016/12/23 0 0 0 0 0
2016/12/26 0.008143306 -0.02443612 0.01037693 -0.002482176 -0.008571446
2016/12/27 -0.004846513 0.003853559 -0.005868793 -0.008709192 -0.002881825
2016/12/28 0.009740233 -0.005758166 -0.007379293 0.000313759 0.005780339
2016/12/29 -0.00803858 -0.01544404 -0.004248083 -0.012233377 -0.011494279