mirror of
https://github.com/mihakralj/QuanTAlib.git
synced 2026-08-07 13:37:44 +00:00
231 lines
29 KiB
Plaintext
231 lines
29 KiB
Plaintext
|
ó
|
|||
|
|
ËI£iùG ã ó² • S r SSKJr SSKrSSKJs Jr SSK r
|
|||
|
|
SSK r SSK
|
|||
|
|
rSSKrSSKJrJrJrJrJrJrJrJrJrJrJrJ r J r J r J r Sr Sr!Sr"Sr#\ S S
|
|||
|
|
S S 4 S!S
|
|||
|
|
jjr$\$" \!5 r%\ RL " \%SS9r'\"\#SS. S"S jjr( " S S5 r) " S S5 r* " S S5 r+ " S S5 r, " S S 5 r- " S S 5 r. " S S 5 r/g)#uÚ test_pandas_ta_parity.py — Validate quantalib vs pandas-ta using the same
|
|||
|
|
methodology as our C# ValidationHelper:
|
|||
|
|
|
|||
|
|
1. Generate a LONG GBM series (5000 bars, seeded) so recursive indicators converge.
|
|||
|
|
2. Compare only the LAST 100 values (DefaultVerificationCount = 100).
|
|||
|
|
3. Skip lookback/warmup bars before comparison window.
|
|||
|
|
4. Tolerance: 1e-7 default, looser for known algorithmic differences (SPEC §9.3).
|
|||
|
|
|
|||
|
|
This mirrors lib/feeds/gbm/ValidationHelper.cs exactly.
|
|||
|
|
é )Ú annotationsN)ÚsmaÚemaÚdemaÚtemaÚwmaÚhmaÚtrimaÚalmaÚrsiÚrocÚmomÚstddevÚvarianceÚzscoreÚbbandsé* é' éd ç•Ö&è .>g Y@gš™™™™™©?gš™™™™™É?gAAp?c óÆ • [ R R U5 nUR U S-
|
|||
|
|
5 nUSUS- - -
|
|||
|
|
U- nU[ R " U5 - U- n X‰- n
|
|||
|
|
[ R
|
|||
|
|
" U [ R S9n X+S' [ R " X«SS S9 U SS=== [ R " U5 -
|
|||
|
|
sss& [ R " U SS U SS S9 X+S' U $ )zsGenerate GBM close prices matching C# GBM feed logic.
|
|||
|
|
|
|||
|
|
S(t+1) = S(t) * exp((mu - sigma^2/2)*dt + sigma*sqrt(dt)*Z)
|
|||
|
|
é g à?é )Údtyper N)Úout)
|
|||
|
|
ÚnpÚrandomÚ default_rngÚstandard_normalÚsqrtÚemptyÚfloat64ÚcumsumÚlogÚexp) ÚnÚseedÚ start_priceÚmuÚsigmaÚdtÚrngÚzÚdriftÚ diffusionÚ log_returnsÚpricess Ú9c:\github\quantalib\python\tests\test_pandas_ta_parity.pyÚ
|
|||
|
|
_generate_gbmr3 # sÏ € ô
|
|||
|
|
�)‰)×
|
|||
|
|
Ñ
|
|||
|
|
Ó
|
|||
|
|
%€CØ ×јA ™EÓ"€AØ
|
|||
|
|
�#˜ ™
|
|||
|
|
Ñ"Ñ
|
|||
|
|
" bÑ (€EØœŸš › Ñ# aÑ'€IØÑ#€KÜ
|
|||
|
|
�XŠX�aœrŸz™zÑ
|
|||
|
|
*€F؈1�I܇I‚Iˆk a b˜zÒ*Ø
|
|||
|
|
ˆ1ˆ2ƒJ”"—&’&˜Ó%Ñ%ƒJ܇F‚Fˆ6�!�"ˆ:˜6 ! "˜:Ò&؈1�IØ €Mó Úclose©ÚnameÚ )Ú verify_countÚ toleranceÚlabelc ó’
|
|||
|
|
• [ U[ R 5 ( a UR 5 OUn[ U 5 n[ U5 ovU:H oˆ( GdÜ [
|
|||
|
|
R " SU4SXg45 S[ R " 5 ; d [
|
|||
|
|
R " [ 5 ( a [
|
|||
|
|
R " [ 5 OSS[ R " 5 ; d [
|
|||
|
|
R " U 5 ( a [
|
|||
|
|
R " U 5 OS[
|
|||
|
|
R " U5 S[ R " 5 ; d [
|
|||
|
|
R " [ 5 ( a [
|
|||
|
|
R " [ 5 OSS[ R " 5 ; d [
|
|||
|
|
R " U5 ( a [
|
|||
|
|
R " U5 OS[
|
|||
|
|
R " U5 S.- n [
|
|||
|
|
R " U S[ U 5 S[ U5 35 S - S
|
|||
|
|
U 0- n
|
|||
|
|
[ [
|
|||
|
|
R " U
|
|||
|
|
5 5 eS =n=p‡[ U 5 n [ S X²-
|
|||
|
|
5 n X S n
|
|||
|
|
X\S n[ R " U
|
|||
|
|
5 [ R " U5 - n[ R" nU" U5 nS oxU:„ nU( Gd> [
|
|||
|
|
R " S
|
|||
|
|
U4SX‡45 S[ R " 5 ; d [
|
|||
|
|
R " [ 5 ( a [
|
|||
|
|
R " [ 5 OS[
|
|||
|
|
R " U5 S[ R " 5 ; d [
|
|||
|
|
R " U5 ( a [
|
|||
|
|
R " U5 OS[
|
|||
|
|
R " U5 [
|
|||
|
|
R " U5 S.- n [
|
|||
|
|
R " U SU 35 S - S
|
|||
|
|
U 0- n
|
|||
|
|
[ [
|
|||
|
|
R " U
|
|||
|
|
5 5 eS =n=n=nnXß nXï n[% [ R " [ R& " UU-
|
|||
|
|
5 5 5 nUU:* nU( Gd
|
|||
|
|
[
|
|||
|
|
R " SU4SUU45 S[ R " 5 ; d [
|
|||
|
|
R " U5 ( a [
|
|||
|
|
R " U5 OSS[ R " 5 ; d [
|
|||
|
|
R " U5 ( a [
|
|||
|
|
R " U5 OSS.- n[
|
|||
|
|
R " U SUS SUS S [ U5 S U S 3
|
|||
|
|
5 S - S U0- n[ [
|
|||
|
|
R " U5 5 eS ng )!zÈCompare only the last `verify_count` values where both are finite.
|
|||
|
|
|
|||
|
|
This matches the C# pattern:
|
|||
|
|
int start = Math.Max(0, count - skip);
|
|||
|
|
for (int i = start; i < count; i++) { ... compare ... }
|
|||
|
|
©ú==)zN%(py3)s
|
|||
|
|
{%(py3)s = %(py0)s(%(py1)s)
|
|||
|
|
} == %(py8)s
|
|||
|
|
{%(py8)s = %(py5)s(%(py6)s)
|
|||
|
|
}ÚlenÚqtl_arrÚpta)Úpy0Úpy1Úpy3Úpy5Úpy6Úpy8z: length mismatch qtl=z vs pta=z
|
|||
|
|
>assert %(py10)sÚpy10Nr )Ú>)zH%(py5)s
|
|||
|
|
{%(py5)s = %(py2)s
|
|||
|
|
{%(py2)s = %(py0)s.sum
|
|||
|
|
}(%(py3)s)
|
|||
|
|
} > %(py8)sr Úfinite)rB Úpy2rD rE rG z: no finite values in last ©ú<=)z%(py0)s <= %(py2)sÚmax_diffr: )rB rK z : max_diff=ú.2ez exceeds tolerance=z.0ez (compared z values in last Ú)z
|
|||
|
|
>assert %(py4)sÚpy4)Ú
|
|||
|
|
isinstanceÚpdÚSeriesÚto_numpyr? Ú
|
|||
|
|
@pytest_arÚ_call_reprcompareÚ @py_builtinsÚlocalsÚ_should_repr_global_nameÚ _safereprÚ_format_assertmsgÚAssertionErrorÚ_format_explanationÚmaxr ÚisfiniteÚsumÚfloatÚabs)r@ Ú
|
|||
|
|
pta_resultr9 r: r; rA Ú @py_assert2Ú @py_assert7Ú @py_assert4Ú @py_format9Ú @py_format11ÚcountÚstartÚq_tailÚp_tailrJ Ú @py_assert1Ú @py_assert6Úq_valsÚp_valsrN Ú @py_format3Ú @py_format5s r2 Ú_verify_last_nrt @ sÖ € ô $.¨j¼"¿)¹)×#DÑ#Dˆ*×
|
|||
|
|
Ñ
|
|||
|
|
Ô
|
|||
|
|
È*€CÜ ˆw‹<ð œ3˜s›8ð ˜8Ñ #÷ ô ïé ôˆ<÷ ò ïé ÷õ ïé ðŒ3÷ ö ïé ðŒ3÷ ó ïé ÷õ ïé ðˆw÷ ö ïé ðˆw÷ ò ïé ðˆ<÷ ñ ïé ÷õ ïé ðœ3÷ ö ïé ðœ3÷ ó ïé ÷õ ïé ð˜s÷ ö ïé ð˜s÷ ò ïé ð˜8÷ õ ïé ðØ ˆ'Ð'¬¨G« ~°X¼cÀ#»h¸ZÐH÷÷ ö ïé ÷÷ ô
|
|||
|
|
�‹L€EÜ ��5Ñ'Ó (€Eà
|
|||
|
|
�Vˆ_€FØ
|
|||
|
|
�ˆ[€Fô �[Š[˜Ó
|
|||
|
|
¤2§;¢;¨vÓ#6Ñ
|
|||
|
|
6€FÜ
|
|||
|
|
�6‰6ÐR‰6�&‹>ÐR˜AÐR˜AÑ ×R×R×RÔRˆ>×R×R×R×R×RÕRŒ2×R×R×RÑRŒ2×R×RÕRˆ6×R×R×R×R×RÔR�&×R×R×RÑR�&×R×RÕRˆ>×R×RÓR˜A×R×R×RÐR % Ð(CÀLÀ>Ð R×R×R×R×R×R×RÓRà
|
|||
|
|
‰^€FØ
|
|||
|
|
‰^€Fä”R—V’VœBŸFšF 6¨F¡?Ó 3Ó4Ó5€HØ �yÑ ÷ õ ïé ôˆ8�y÷ ò ïé ÷õ ïé ðˆ8÷ ö ïé ðˆ8÷ ó ïé ÷õ ïé ð�y÷ ö ïé ð�y÷ ÷ ïé ðØ ˆ'�˜X c˜NÐ*=¸iȸ_ð Mܘ“[�MÐ!1°,° ¸qð B÷÷ ö ïé ÷õ r4 c óº • \ rS rSrSrS
|
|||
|
|
S jrS
|
|||
|
|
S jrS
|
|||
|
|
S jr\R R SS9S
|
|||
|
|
S j5 r \R R S S9S
|
|||
|
|
S
|
|||
|
|
j5 r S r
|
|||
|
|
g )Ú TestTrendFIRék uM FIR indicators: SMA, WMA, HMA — deterministic convolution, tight tolerance.c ó‚ • S H9 n[ [ US9n[ R " [ US9n[ X#SSU S3S9 M; g )N©é
|
|||
|
|
é é2 ©Úlengthr zSMA(rP ©r: r; )r ÚCLOSEÚtaÚSERIESrt ©Úselfr~ ÚqtlrA s r2 Útest_smaÚTestTrendFIR.test_sman ó@ € Û"ˆFÜ”e FÑ+ˆCÜ—&’&œ¨Ñ/ˆCÜ ˜3¨tØ#'¨ x¨qÐ!1ô
|
|||
|
|
3ò #r4 c ó‚ • S H9 n[ [ US9n[ R " [ US9n[ X#SSU S3S9 M; g )N©rz é é r} r zWMA(rP r )r r€ r� r‚ rt rƒ s r2 Útest_wmaÚTestTrendFIR.test_wmau rˆ r4 c ó‚ • S H9 n[ [ US9n[ R " [ US9n[ X#SSU S3S9 M; g )N)é r‹ r{ r} ç:Œ0âŽyE>zHMA(rP r )r r€ r� r‚ rt rƒ s r2 Útest_hmaÚTestTrendFIR.test_hma| ó@ € Û!ˆFÜ”e FÑ+ˆCÜ—&’&œ¨Ñ/ˆCÜ ˜3¨tØ#'¨ x¨qÐ!1ô
|
|||
|
|
3ò "r4 u• TRIMA kernel differs: quantalib uses symmetric triangular convolution, pandas-ta delegates to TA-Lib which uses cascaded SMA (SPEC §9.3 known delta)©Úreasonc óh • [ [ SS9n[ R " [ SS9n[ XSSS9 g )Nr‹ r} r z TRIMA(14)r )r
|
|||
|
|
r€ r� r‚ rt ©r„ r… rA s r2 Ú
|
|||
|
|
test_trimaÚTestTrendFIR.test_trimaƒ s, € ô ”E "Ñ%ˆÜ�hŠh”v bÑ)ˆÜ�s¨4°{ÓCr4 uZ ALMA sigma/offset defaults differ between quantalib and pandas-ta (SPEC §9.3 known delta)c óh • [ [ SS9n[ R " [ SS9n[ XSSS9 g )Nr‹ r} ç�íµ ÷Æ°>zALMA(14)r )r r€ r� r‚ rt r˜ s r2 Ú test_almaÚTestTrendFIR.test_alma‹ s, € ô ”5 Ñ$ˆÜ�gŠg”f RÑ(ˆÜ�s¨4°zÓBr4 © N©ÚreturnÚNone)Ú__name__Ú
|
|||
|
|
__module__Ú __qualname__Ú__firstlineno__Ú__doc__r† r� r’ ÚpytestÚmarkÚxfailr™ r� Ú__static_attributes__rŸ r4 r2 rv rv k st † ÙWô3ô3ô3ð ‡[�[×Ñð RÐð SóDóSðDð
|
|||
|
|
‡[�[×Ñð RÐð SóCóSóCr4 rv c ó6 • \ rS rSrSrSS jrSS jrSS jrSrg) Ú TestTrendIIRé— u IIR indicators: EMA, DEMA, TEMA — recursive convergence.
|
|||
|
|
|
|||
|
|
With 5000 bars the warmup difference is buried in the past.
|
|||
|
|
The last 100 bars should match tightly.
|
|||
|
|
c ó‚ • S H9 n[ [ US9n[ R " [ US9n[ X#SSU S3S9 M; g )Nry r} çH¯¼šò×z>zEMA(rP r )r r€ r� r‚ rt rƒ s r2 Útest_emaÚTestTrendIIR.test_emaž rˆ r4 c ó‚ • S H9 n[ [ US9n[ R " [ US9n[ X#SSU S3S9 M; g )Nry r} r° zDEMA(rP r )r r€ r� r‚ rt rƒ s r2 Ú test_demaÚTestTrendIIR.test_dema¥ ó@ € Û"ˆFÜ”u VÑ,ˆCÜ—'’'œ&¨Ñ0ˆCÜ ˜3¨tØ#(¨¨°Ð!2ô
|
|||
|
|
4ò #r4 c ó‚ • S H9 n[ [ US9n[ R " [ US9n[ X#SSU S3S9 M; g )NrŠ r} r° zTEMA(rP r )r r€ r� r‚ rt rƒ s r2 Ú test_temaÚTestTrendIIR.test_tema¬ r¶ r4 rŸ Nr ) r£ r¤ r¥ r¦ r§ r± r´ r¸ r« rŸ r4 r2 r r — s † ñô 3ô4÷4r4 r c ón • \ rS rSrSrS
|
|||
|
|
S jr\R R SS9S
|
|||
|
|
S j5 r S
|
|||
|
|
S jr
|
|||
|
|
Sr g ) Ú TestMomentumé¸ z$Momentum: RSI (recursive), ROC, MOM.c ó‚ • S H9 n[ [ US9n[ R " [ US9n[ X#SSU S3S9 M; g)zqRSI is recursive (Wilder smoothing). With 5000 bars, warmup
|
|||
|
|
convergence difference is negligible in the last 100.)é r‹ é r} rœ zRSI(rP r N)r r€ r� r‚ rt rƒ s r2 Útest_rsiÚTestMomentum.test_rsi» sB € ó "ˆFÜ”e FÑ+ˆCÜ—&’&œ¨Ñ/ˆCÜ ˜3¨tØ#'¨ x¨qÐ!1ô
|
|||
|
|
3ò "r4 u“ ROC formula differs: quantalib uses absolute difference (close-prev), pandas-ta uses percentage ((close/prev - 1)*100). Known delta per SPEC §9.3.r• c óh • [ [ SS9n[ R " [ SS9n[ XSSS9 g)z¶ROC: quantalib 'Roc' is Rate of Change (Absolute) = close - close[n].
|
|||
|
|
pandas-ta 'roc' is Rate of Change (Percentage) = ((c/c[n])-1)*100.
|
|||
|
|
These are fundamentally different indicators.rz r} r° zROC(10)r N)r
|
|||
|
|
r€ r� r‚ rt r˜ s r2 Útest_rocÚTestMomentum.test_rocÄ s, € ô ”% Ñ#ˆÜ�fŠf”V BÑ'ˆÜ�s¨4°yÓAr4 c ó‚ • S H9 n[ [ US9n[ R " [ US9n[ X#SSU S3S9 M; g )N)é rz r{ r} r zMOM(rP r )r r€ r� r‚ rt rƒ s r2 Útest_momÚTestMomentum.test_momÐ r” r4 rŸ Nr ) r£ r¤ r¥ r¦ r§ rÀ r¨ r© rª rà rÇ r« rŸ r4 r2 r» r» ¸ s< † Ù.ô3ð ‡[�[×Ñð =Ðð >óBó >ðB÷3r4 r» c ó6 • \ rS rSrSrSS jrSS jrSS jrSrg) ÚTestStatisticséÜ zÓSTDDEV, VARIANCE, ZSCORE.
|
|||
|
|
|
|||
|
|
Note: pandas-ta uses sample stddev (ddof=1), quantalib may use population.
|
|||
|
|
With 5000 bars and period=20, the difference is ~5% for ddof effect.
|
|||
|
|
We use relative tolerance where needed.
|
|||
|
|
c óØ • S GHã n[ [ US9n[ R " [ US9nUR 5 n[
|
|||
|
|
U5 n[ SU[ -
|
|||
|
|
5 nX&S nXFS n[ R " U5 [ R " U5 - n [ R " U 5 S:X a [ R " SU S35 Xy n
|
|||
|
|
X‰ n [ [ R " [ R " X«-
|
|||
|
|
5 5 5 n U S:: a g U
|
|||
|
|
[ R " XS-
|
|||
|
|
- 5 - n
|
|||
|
|
[ [ R " [ R " XÛ-
|
|||
|
|
5 5 5 nUS:: a g U
|
|||
|
|
[ R " US-
|
|||
|
|
U- 5 - n[ [ R " [ R " Xû-
|
|||
|
|
5 5 5 nUS:: a g [ R " SU SU S S
|
|||
|
|
US S US 35 GMæ g ) Nry r} r zSTDDEV(ú): no finite in tailr‘ r ú
|
|||
|
|
): direct=rO õ , pop→sample=õ , sample→pop=)r r€ r� Ústdevr‚ rU r? r_ Ú VERIFY_COUNTr r` ra r¨ Úfailrb rc r ©r„ r~ r… rA Úpta_nprj rk rl rm rJ Úq_fÚp_frN ÚadjustedÚadj_diffÚ adjusted_invÚinv_diffs r2 Ú test_stddevÚTestStatistics.test_stddevä sŸ € Ü"ˆFÜœ vÑ.ˆCÜ—(’(œ6¨&Ñ1ˆCð —\‘\“^ˆFܘ“HˆEܘ˜5¤<Ñ/Ó0ˆEؘ�[ˆFؘF�^ˆFÜ—[’[ Ó(¬2¯;ª;°vÓ+>Ñ>ˆFä�vŠv�f‹~ Ó"Ü— ’ ˜g f XÐ-AÐ BÔCà‘.ˆCØ‘.ˆCô œRŸVšV¤B§F¢F¨3©9Ó$5Ó 6Ó7ˆHؘ4Ó Ùð œRŸWšW V¸©zÑ%:Ó ;Ñ;ˆHÜ œRŸVšV¤B§F¢F¨8©>Ó$:Ó ;Ó<ˆHؘ4Ó Ù𠤧¢¨&°1©*¸Ñ)>Ó!?Ñ?ˆLÜ œRŸVšV¤B§F¢F¨<Ñ+=Ó$>Ó ?Ó@ˆHؘ4Ó Ùä �KŠKؘ&˜ ¨H°S¨>ð : Ø (¨˜~¨_¸XÀc¸NðL÷
|
|||
|
|
òK #r4 c ó† • S GHº n[ [ US9n[ R " [ US9nUR 5 n[ U5 n[
|
|||
|
|
SU[ -
|
|||
|
|
5 nX&S nXFS n[ R " U5 [ R " U5 - n [ R " U 5 S:X a [ R " SU S35 Xy n
|
|||
|
|
X‰ n [ [ R " [ R " X«-
|
|||
|
|
5 5 5 n U S:: a g X¡US-
|
|||
|
|
- - n
|
|||
|
|
[ [ R " [ R " XÛ-
|
|||
|
|
5 5 5 nUS:: a g X¡S-
|
|||
|
|
U- - n[ [ R " [ R " Xû-
|
|||
|
|
5 5 5 nUS:: a g [ R " SU SU S S
|
|||
|
|
US S US 35 GM½ g ) Nry r} r z VARIANCE(rÍ r‘ r rÎ rO rÏ rÐ )r r€ r� r‚ rU r? r_ rÒ r r` ra r¨ rÓ rb rc rÔ s r2 Ú
|
|||
|
|
test_varianceÚ TestStatistics.test_variance s‡ € Ü"ˆFÜœ5¨Ñ0ˆCÜ—+’+œf¨VÑ4ˆCØ—\‘\“^ˆFܘ“HˆEܘ˜5¤<Ñ/Ó0ˆEؘ�[ˆFؘF�^ˆFÜ—[’[ Ó(¬2¯;ª;°vÓ+>Ñ>ˆFä�vŠv�f‹~ Ó"Ü— ’ ˜i¨ xÐ/CÐ DÔEà‘.ˆCØ‘.ˆCô œRŸVšV¤B§F¢F¨3©9Ó$5Ó 6Ó7ˆHؘ4Ó Ùð ¨°©
|
|||
|
|
Ñ 3Ñ4ˆHÜ œRŸVšV¤B§F¢F¨8©>Ó$:Ó ;Ó<ˆHؘ4Ó Ùà ¨A¡:°Ñ"7Ñ8ˆLÜ œRŸVšV¤B§F¢F¨<Ñ+=Ó$>Ó ?Ó@ˆHؘ4Ó Ùä �KŠKؘF˜8 :¨h°s¨^ð < Ø (¨˜~¨_¸XÀc¸NðL÷
|
|||
|
|
òA #r4 c ó • S GHy n[ [ US9n[ R " [ US9nUR 5 n[ U5 n[
|
|||
|
|
SU[ -
|
|||
|
|
5 nX&S nXFS n[ R " U5 [ R " U5 - n [ R " U 5 S:X a [ R " SU S35 Xy n
|
|||
|
|
X‰ n [ R " XS-
|
|||
|
|
- 5 n [ [ R " [ R " X«-
|
|||
|
|
5 5 5 n
|
|||
|
|
[ [ R " [ R " X«U - -
|
|||
|
|
5 5 5 n[ [ R " [ R " X¬- U -
|
|||
|
|
5 5 5 n[! XÞU5 nSnUU:* nU( dÑ [" R$ " S U4S
|
|||
|
|
UU45 S [& R( " 5 ; d [" R* " U5 ( a [" R, " U5 OS [" R, " U5 S .- n[" R. " SU S
|
|||
|
|
US SU
|
|||
|
|
S SUS SUS S3 5 S- SU0- n[1 [" R2 " U5 5 eS=nnGM| g)z¶ZSCORE = (x - mean) / stddev.
|
|||
|
|
|
|||
|
|
quantalib uses population stddev (ddof=0), pandas-ta uses sample (ddof=1).
|
|||
|
|
The ratio is sqrt(n/(n-1)). We verify after applying the correction factor.
|
|||
|
|
ry r} r NzZSCORE(rÍ r r° rL )z%(py0)s <= %(py3)sÚbest)rB rD z
|
|||
|
|
): best_diff=rO z (direct=z , corrected=z, inv=rP z
|
|||
|
|
>assert %(py5)srE )r r€ r� r‚ rU r? r_ rÒ r r` ra r¨ rÓ r rb rc ÚminrV rW rX rY rZ r[ r\ r] r^ )r„ r~ r… rA rÕ rj rk rl rm rJ rÖ r× Ú
|
|||
|
|
ddof_ratioÚ diff_directÚdiff_correctedÚdiff_invrâ re rn Ú @py_format4Ú @py_format6s r2 Ú test_zscoreÚTestStatistics.test_zscore5 s% € ô #ˆFÜœ vÑ.ˆCÜ—)’)œF¨6Ñ2ˆCØ—\‘\“^ˆFܘ“HˆEܘ˜5¤<Ñ/Ó0ˆEؘ�[ˆFؘF�^ˆFÜ—[’[ Ó(¬2¯;ª;°vÓ+>Ñ>ˆFä�vŠv�f‹~ Ó"Ü— ’ ˜g f XÐ-AÐ BÔCà‘.ˆCØ‘.ˆCô Ÿš °A©:Ñ!6Ó7ˆJô ¤§¢¤r§v¢v¨c©iÓ'8Ó 9Ó:ˆKÜ "¤2§6¢6¬"¯&ª&°¸ZÑ7GÑ1GÓ*HÓ#IÓ JˆNÜ œRŸVšV¤B§F¢F¨3Ñ+;¸cÑ+AÓ$BÓ CÓDˆHä�{°HÓ=ˆDØ ð
|
|||
|
|
�4˜4‘<÷
|
|||
|
|
ô
|
|||
|
|
ïé ô
|
|||
|
|
�4˜4÷
|
|||
|
|
ò
|
|||
|
|
÷ �v÷
|
|||
|
|
õ
|
|||
|
|
ïé ð
|
|||
|
|
�4÷
|
|||
|
|
ö
|
|||
|
|
÷ �Ið
|
|||
|
|
�4÷
|
|||
|
|
ò
|
|||
|
|
÷ �Ið
|
|||
|
|
˜4÷
|
|||
|
|
õ
|
|||
|
|
ïé ð
|
|||
|
|
ؘ&˜ ¨t°C¨jð 9Ø& sÐ+¨<¸ÀsÐ7Kð LØ �n Að'÷
|
|||
|
|
÷
|
|||
|
|
ö
|
|||
|
|
ïé ÷
|
|||
|
|
÷
|
|||
|
|
ñ
|
|||
|
|
ò; #r4 rŸ Nr ) r£ r¤ r¥ r¦ r§ rÜ rß rê r« rŸ r4 r2 rÊ rÊ Ü s † ñô)ôV$÷L'r4 rÊ c óD • \ rS rSrSrS S
|
|||
|
|
S jjrS S jrS S jrS S jrSr g) ÚTestMultiOutputic zEMulti-output indicators: BBands returns (upper, middle, lower) tuple.c ó| • [ [ XS9n[ R " [ XSS9n[ U[
|
|||
|
|
5 ( a US US US pvnO^[
|
|||
|
|
US5 ( a+ UR S:X a USS2S4 USS2S4 USS2S4 pvnO"[ R " S[ U5 35 [ UR 5 nU V s/ s H o™R S 5 ( d M U PM n
|
|||
|
|
n U V s/ s H o™R S
|
|||
|
|
5 ( d M U PM n n U V s/ s H o™R S 5 ( d M U PM n n / oÚoê( a
|
|||
|
|
U oë( a U nU( Gd� S S
|
|||
|
|
S[ R " 5 ; d [ R" " U
|
|||
|
|
5 ( a [ R$ " U
|
|||
|
|
5 OS0- oýR' U5 U
|
|||
|
|
( aË SSS[ R " 5 ; d [ R" " U 5 ( a [ R$ " U 5 OS0- nU
|
|||
|
|
R' U5 U ( ab SSS[ R " 5 ; d [ R" " U 5 ( a [ R$ " U 5 OS0- nU
|
|||
|
|
R' U5 [ R( " U
|
|||
|
|
S5 0 - n[ R* " SU 35 S- SU0- n[- [ R. " U5 5 eS=píXJS R1 5 nXKS R1 5 nXLS R1 5 nWWW4UUU44$ s sn f s sn f s sn f )z0Get both quantalib and pandas-ta BBands results.)r~ Ústdr )r~ rï Úddofr r ÚndimNz Unexpected bbands return type: ÚBBUÚBBMÚBBLz%(py2)srK Úbbuz%(py4)srQ Úbbmz%(py6)srF ÚbblzBBands columns not found: z
|
|||
|
|
>assert %(py9)sÚpy9)r r€ r� r‚ rR ÚtupleÚhasattrrñ r¨ rÓ ÚtypeÚlistÚcolumnsÚ
|
|||
|
|
startswithrX rY rV rZ r[ ÚappendÚ_format_boolopr\ r] r^ rU )r„ r~ rï r… rA Ú qtl_upperÚqtl_midÚ qtl_lowerÚcolsÚcrõ rö r÷ rn Ú @py_assert0rr rs Ú @py_format7Ú @py_format8Ú @py_format10Ú pta_upperÚpta_midÚ pta_lowers r2 Ú _get_bbandsÚTestMultiOutput._get_bbandsf s- € ä”U 6Ñ3ˆä�iŠiœ v¸QÑ?ˆô �cœ5× !Ñ !Ø,/°©F°C¸±F¸CÀ¹F ˆI Ü
|
|||
|
|
�S˜&×
|
|||
|
|
!Ñ
|
|||
|
|
! c§h¡h°!£mØ,/²°1°©I°sº1¸a¸4±yÀ#ÂaÈÀdÁ) ˆI ä �KŠKÐ9¼$¸s»)¸ÐEÔ Fô
|
|||
|
|
�C—K‘KÓ ˆÙ Ó6š$�Q§,¡,¨u×"5�q™$ˆÐ6Ù Ó6š$�Q§,¡,¨u×"5�q™$ˆÐ6Ù Ó6š$�Q§,¡,¨u×"5�q™$ˆÐ6ÚG�s�sÐG•s˜s×G×G×G×G×G×GÓGˆs×G×G×GÑGˆs×G×G×GÑGŽs×G×G×G×G×GÒG�s×G×G×GÑG�s×G×G×GÒG–s×G×G×G×G×GÒG˜s×G×G×GÑG˜s×G×G×G×G×G×G×G×GÔGÐ&@ÀÀÐ$G×G×G×G×G×GÖGà˜A™‘K×(Ñ(Ó*ˆ ؘ!‘f‘+×&Ñ&Ó(ˆØ˜A™‘K×(Ñ(Ó*ˆ à˜7 IÐ.°¸GÀYÐ0OÐOÐOùò 7ùÚ6ùÚ6s$ ÃL/Ã!L/Ã-L4Ä
|
|||
|
|
L4ÄL9Ä3L9c óL • U R 5 u u pnu pn[ X#SSS9 g)z(Middle band = SMA, should match exactly.r z
|
|||
|
|
BBands middler N©r
|