通常のデバッグ
たまに小型株とか買うと
このように銘柄のデータがないよと怒られます。
よくあるデバッグの方法として
https://qiita.com/investor-gon/items/64acc909257fa3448e27
ctx.loggger.debug(cp)
と株価の終値をdebugするのが一般的ですが、、
問題が
QuantXのcpですがこのようにDataFrameの形でdebugされます。
しかし、銘柄が多かったり期間が長かったりするとログが省略されます。
jp.stock.2914 jp.stock.3092 ... jp.stock.5302 \
date ...
2014-01-06 3425.0 848.666687 ... 2000.0
2014-01-07 3375.0 816.666687 ... 2010.0
2014-01-08 3290.0 814.000000 ... 2030.0
2014-01-09 3180.0 833.666687 ... 2010.0
2014-01-10 3195.0 844.000000 ... 2000.0
2014-01-14 3172.0 813.666687 ... 1970.0
2014-01-15 3221.0 833.333313 ... 2010.0
2014-01-16 3220.0 822.333313 ... 2020.0
2014-01-17 3186.0 845.666687 ... 2090.0
2014-01-20 3200.0 813.666687 ... 2110.0
2014-01-21 3231.0 827.666687 ... 2080.0
2014-01-22 3235.0 831.333313 ... 2060.0
2014-01-23 3227.0 840.000000 ... 2020.0
2014-01-24 3173.0 796.000000 ... 1990.0
2014-01-27 3128.0 756.666687 ... 1910.0
2014-01-28 3068.0 763.333313 ... 1900.0
2014-01-29 3129.0 787.000000 ... 1950.0
2014-01-30 3026.0 752.333313 ... 1900.0
2014-01-31 3197.0 763.333313 ... 1890.0
2014-02-03 3135.0 764.333313 ... 1820.0
2014-02-04 3070.0 730.000000 ... 1700.0
2014-02-05 3138.0 749.666687 ... 1730.0
2014-02-06 3107.0 719.000000 ... 1760.0
2014-02-07 3186.0 731.000000 ... 1820.0
2014-02-10 3285.0 773.000000 ... 1810.0
2014-02-12 3347.0 750.000000 ... 1920.0
2014-02-13 3319.0 726.333313 ... 1890.0
2014-02-14 3280.0 701.666687 ... 1840.0
2014-02-17 3291.0 717.666687 ... 1850.0
2014-02-18 3357.0 736.333313 ... 1910.0
... ... ... ... ...
2017-02-17 3768.0 2452.000000 ... 2980.0
2017-02-20 3774.0 2487.000000 ... 2990.0
2017-02-21 3778.0 2487.000000 ... 2980.0
2017-02-22 3798.0 2456.000000 ... 3000.0
2017-02-23 3811.0 2384.000000 ... 2960.0
2017-02-24 3805.0 2367.000000 ... 2910.0
2017-02-27 3778.0 2346.000000 ... 2900.0
2017-02-28 3757.0 2348.000000 ... 2890.0
2017-03-01 3800.0 2418.000000 ... 2910.0
2017-03-02 3776.0 2421.000000 ... 2930.0
2017-03-03 3781.0 2409.000000 ... 2920.0
2017-03-06 3782.0 2402.000000 ... 2970.0
2017-03-07 3780.0 2378.000000 ... 2970.0
2017-03-08 3766.0 2361.000000 ... 2960.0
2017-03-09 3755.0 2361.000000 ... 3000.0
2017-03-10 3804.0 2406.000000 ... 3000.0
2017-03-13 3845.0 2407.000000 ... 3000.0
2017-03-14 3840.0 2435.000000 ... 2990.0
2017-03-15 3831.0 2436.000000 ... 2940.0
2017-03-16 3783.0 2385.000000 ... 3200.0
2017-03-17 3772.0 2373.000000 ... 3220.0
2017-03-21 3856.0 2330.000000 ... 3170.0
2017-03-22 3752.0 2293.000000 ... 3070.0
2017-03-23 3753.0 2284.000000 ... 3080.0
2017-03-24 3786.0 2302.000000 ... 3220.0
2017-03-27 3725.0 2321.000000 ... 3180.0
2017-03-28 3726.0 2382.000000 ... 3200.0
2017-03-29 3701.0 2422.000000 ... 3360.0
2017-03-30 3654.0 2441.000000 ... 3310.0
2017-03-31 3618.0 2463.000000 ... 3340.0
そもそも長い
解決法
NaNをカウントすればいいのでは？
NaN_count = cp.isnull().sum()
ctx.logger.debug(NaN_count)
結果
jp.stock.2121 0
jp.stock.2124 0
jp.stock.2337 0
jp.stock.2352 0
jp.stock.2371 0
jp.stock.2379 0
jp.stock.2391 0
jp.stock.2477 0
jp.stock.2914 0
jp.stock.3092 0
jp.stock.3543 608
jp.stock.3647 0
jp.stock.3665 0
jp.stock.3679 0
jp.stock.3712 0
jp.stock.3723 0
jp.stock.3750 0
jp.stock.3763 0
jp.stock.3772 0
jp.stock.3835 0
jp.stock.3848 0
jp.stock.3901 234
jp.stock.3922 548
jp.stock.3963 670
jp.stock.3969 727
jp.stock.3979 783
jp.stock.3997 794
jp.stock.4327 0
jp.stock.4368 0
jp.stock.4369 0
jp.stock.4507 0
jp.stock.4521 0
jp.stock.4704 0
jp.stock.4732 0
jp.stock.4792 0
jp.stock.4820 0
jp.stock.4849 0
jp.stock.4975 0
jp.stock.5194 0
jp.stock.5301 0
jp.stock.5302 0
jp.stock.6048 324
jp.stock.6080 0
jp.stock.6146 0
jp.stock.6157 0
jp.stock.6161 0
jp.stock.6196 603
jp.stock.6538 721
jp.stock.6539 722
jp.stock.6552 794
このようにNaNがあるものは数字が帰ってきます。