change significance level

[yeestrada]

Hi everyone, I am newbie in R, rigth now I am trying to use ks.test, ad.test and chisq.test from goftest, but i can not find the way to change the significance level of those tests. Actually I am using them with the rpy2 library, because I am currently working with python and calling R functions using that library, everything is just fine except, I can't find how to change the significance level. can anyone help me?

[GreyMerchant]

It is very difficult to help you with only that. I suggest make a simple reprex and we'll be able to help (See here: FAQ: How to do a minimal reproducible example ( reprex ) for beginners)

[yeestrada]

Hi, sorry here is part of my code:

import rpy2.robjects as R
import rpy2.robjects.packages as R_packages

R_test, R_mass = R.packages.importr('goftest'), R.packages.importr('MASS')

R_ksr_test, R_ad_test, R_chisq_test = R.r['ks.test'], R.r['ad.test'], R.r['chisq.test']
distributions = [('norm', 'normal'), # mean, deviation
('pexp', 'exponential'), # rate
('pweibull', 'weibull') # shape, scale
]

def make_test(data, distribution, conf, test='ad'):
args = {}
dist, name = distribution
if name in mean_param: args['mean'] = conf['mean']
if name in sd_param: args['sd'] = conf['sd']
if name in shape_param: args['shape'] = conf['shape']
if name in scale_param: args['scale'] = conf['scale']
if name in rate_param: args['rate'] = conf['rate']

   if test == 'ks':
       return R_ksr_test(data, dist, **args)
   elif test == 'chi2':
       return R_chisq_test(data, dist, **args)
   return R_ad_test(data, dist, **args)

def get_fitdistr(items, distribution):
data = {}
response = R_mass.fitdistr(items, distribution)
if distribution == 'weibull':
data['shape'], data['scale'] = response[0]
elif distribution == 'exponential':
data['rate'] = response[0][0]
else:
data['mean'], data['sd'] = response[0]
return data

vector = [9.2, 10.5, 10.8, 12.3, 14.8, 15.6, 16.1, 19.5, 22.4, 24.7, 25.3, 26.5, 29.9, 30.3, 30.7, 35.6, 46.5, 50.4, 68.2, 90.6]
vector = R.IntVector(vector)
data = u.get_fitdistr(vector, distributions[2][1])
stat, p_val, method, name = u.make_test(vector, distributions[2], data)

As you can see I am calling ad.test, ks.test and chisq.test without problems, but I need to offer the choice of change the significance level in each test.