Module stikpetP.other.thumb_rank_biserial
Expand source code
import pandas as pd
from ..effect_sizes.convert_es import es_convert
from ..other.thumb_cohen_d import th_cohen_d
from ..other.thumb_gk_gamma import th_gk_gamma
from ..other.thumb_somers_d import th_somers_d
from ..other.thumb_cliff_delta import th_cliff_delta
def th_rank_biserial(rb, version="glass", qual=None, convert="no"):
'''
Rule of thumb for Rank Biserial Correlation
--------------------------
Simple function to use a rule-of-thumb for the Rank Biserial Correlation.
This function is shown in this [YouTube video](https://youtu.be/Tx4wJxuh5AM) and the measure is also described at [PeterStatistics.com](https://peterstatistics.com/Terms/Correlations/RankBiserialCorrelation.html)
Parameters
----------
rb : float
the rank-biserial correlation value
version : {"glass", "cureton"}, optional
version of rank-biserial that was used
qual : see notes, optional
indication which set of rule-of-thumb to use.
convert : {"no", "cohen_d", "vda"}, optional
conversion to use (only for Glass version)
Returns
-------
pandas.DataFrame
A dataframe with the following columns:
* *classification*, the qualification of the effect size
* *reference*, a reference for the rule of thumb used
Notes
-----
If a Cureton version of rank-biserial was used, the result for independent samples is the same as Goodman-Kruskal gamma, so we can also use those rules-of-thumb.
If a Glass version is used, then we can either use Cliff Delta, Somers d, or a conversion to either Cohen d, or Vargha-Delaney A.
See the separate functions on each of these for various rules-of-thumbs.
Before, After and Alternatives
------------------------------
Before this you might want to obtain the measure:
* [r_rank_biserial_os](../correlations/cor_rank_biserial_os.html#r_rank_biserial_os) to determine a the rank biserial for one-sample
* [r_rank_biserial_is](../correlations/r_rank_biserial_is.html#r_rank_biserial_is) to determine a the rank biserial for independent samples
The function uses the convert function and corresponding rules of thumb:
* [es_convert](../effect_sizes/convert_es.html#es_convert) for the conversions
Author
------
Made by P. Stikker
Companion website: https://PeterStatistics.com
YouTube channel: https://www.youtube.com/stikpet
Donations: https://www.patreon.com/bePatron?u=19398076
'''
if version == 'cureton':
if qual is None:
qual = "blaikie"
results = th_gk_gamma(rb, qual)
elif version=='glass':
if convert == "cohen_d":
if qual is None:
qual = "sawilowsky"
d = es_convert(rb, "rb", "cohend")
results = th_cohen_d(d, qual)
elif convert == "vda" or convert == "cles":
if qual is None:
qual = "vargha"
vda = es_convert(rb, "rb", "cle")
results = th_cle(vda, qual)
else:
if qual == "metsamuuronen-somers":
results = th_somers_d(rb, "metsamuuronen")
else:
if qual is None:
qual = "romano"
results = th_cliff_delta(rb, qual)
return(results)Functions
def th_rank_biserial(rb, version='glass', qual=None, convert='no')-
Rule Of Thumb For Rank Biserial Correlation
Simple function to use a rule-of-thumb for the Rank Biserial Correlation.
This function is shown in this YouTube video and the measure is also described at PeterStatistics.com
Parameters
rb:float- the rank-biserial correlation value
version:{"glass", "cureton"}, optional- version of rank-biserial that was used
qual:see notes, optional- indication which set of rule-of-thumb to use.
convert:{"no", "cohen_d", "vda"}, optional- conversion to use (only for Glass version)
Returns
pandas.DataFrame-
A dataframe with the following columns:
- classification, the qualification of the effect size
- reference, a reference for the rule of thumb used
Notes
If a Cureton version of rank-biserial was used, the result for independent samples is the same as Goodman-Kruskal gamma, so we can also use those rules-of-thumb.
If a Glass version is used, then we can either use Cliff Delta, Somers d, or a conversion to either Cohen d, or Vargha-Delaney A.
See the separate functions on each of these for various rules-of-thumbs.
Before, After and Alternatives
Before this you might want to obtain the measure: * r_rank_biserial_os to determine a the rank biserial for one-sample * r_rank_biserial_is to determine a the rank biserial for independent samples
The function uses the convert function and corresponding rules of thumb: * es_convert for the conversions
Author
Made by P. Stikker
Companion website: https://PeterStatistics.com
YouTube channel: https://www.youtube.com/stikpet
Donations: https://www.patreon.com/bePatron?u=19398076Expand source code
def th_rank_biserial(rb, version="glass", qual=None, convert="no"): ''' Rule of thumb for Rank Biserial Correlation -------------------------- Simple function to use a rule-of-thumb for the Rank Biserial Correlation. This function is shown in this [YouTube video](https://youtu.be/Tx4wJxuh5AM) and the measure is also described at [PeterStatistics.com](https://peterstatistics.com/Terms/Correlations/RankBiserialCorrelation.html) Parameters ---------- rb : float the rank-biserial correlation value version : {"glass", "cureton"}, optional version of rank-biserial that was used qual : see notes, optional indication which set of rule-of-thumb to use. convert : {"no", "cohen_d", "vda"}, optional conversion to use (only for Glass version) Returns ------- pandas.DataFrame A dataframe with the following columns: * *classification*, the qualification of the effect size * *reference*, a reference for the rule of thumb used Notes ----- If a Cureton version of rank-biserial was used, the result for independent samples is the same as Goodman-Kruskal gamma, so we can also use those rules-of-thumb. If a Glass version is used, then we can either use Cliff Delta, Somers d, or a conversion to either Cohen d, or Vargha-Delaney A. See the separate functions on each of these for various rules-of-thumbs. Before, After and Alternatives ------------------------------ Before this you might want to obtain the measure: * [r_rank_biserial_os](../correlations/cor_rank_biserial_os.html#r_rank_biserial_os) to determine a the rank biserial for one-sample * [r_rank_biserial_is](../correlations/r_rank_biserial_is.html#r_rank_biserial_is) to determine a the rank biserial for independent samples The function uses the convert function and corresponding rules of thumb: * [es_convert](../effect_sizes/convert_es.html#es_convert) for the conversions Author ------ Made by P. Stikker Companion website: https://PeterStatistics.com YouTube channel: https://www.youtube.com/stikpet Donations: https://www.patreon.com/bePatron?u=19398076 ''' if version == 'cureton': if qual is None: qual = "blaikie" results = th_gk_gamma(rb, qual) elif version=='glass': if convert == "cohen_d": if qual is None: qual = "sawilowsky" d = es_convert(rb, "rb", "cohend") results = th_cohen_d(d, qual) elif convert == "vda" or convert == "cles": if qual is None: qual = "vargha" vda = es_convert(rb, "rb", "cle") results = th_cle(vda, qual) else: if qual == "metsamuuronen-somers": results = th_somers_d(rb, "metsamuuronen") else: if qual is None: qual = "romano" results = th_cliff_delta(rb, qual) return(results)