def check_internet():
"""Check if internet is available"""
try:
urlopen("https://github.com")
except URLError as err:
return False
return True
def check_internet():
"""Check if internet is available"""
try:
urlopen("https://github.com")
except URLError as err:
return False
return True
def check_internet(url=None):
"""Check if internet is available"""
url = "https://github.com" if url is None else url
try:
urlopen(url)
except URLError as err:
return False
return True
def webuse(data, baseurl='http://www.stata-press.com/data/r11/', as_df=True):
"""
Parameters
----------
data : str
Name of dataset to fetch.
baseurl : str
The base URL to the stata datasets.
as_df : bool
If True, returns a `pandas.DataFrame`
Returns
-------
dta : Record Array
A record array containing the Stata dataset.
Examples
--------
>>> dta = webuse('auto')
Notes
-----
Make sure baseurl has trailing forward slash. Doesn't do any
error checking in response URLs.
"""
# lazy imports
from statsmodels.iolib import genfromdta
url = urljoin(baseurl, data+'.dta')
dta = urlopen(url)
dta = StringIO(dta.read()) # make it truly file-like
if as_df: # could make this faster if we don't process dta twice?
return DataFrame.from_records(genfromdta(dta))
else:
return genfromdta(dta)
def _urlopen_cached(url, cache):
"""
Tries to load data from cache location otherwise downloads it. If it
downloads the data and cache is not None then it will put the downloaded
data in the cache path.
"""
from_cache = False
if cache is not None:
file_name = url.split("://")[-1].replace('/', ',')
file_name = file_name.split('.')
if len(file_name) > 1:
file_name[-2] += '-v2'
else:
file_name[0] += '-v2'
file_name = '.'.join(file_name) + ".zip"
cache_path = join(cache, file_name)
try:
data = _open_cache(cache_path)
from_cache = True
except:
pass
# not using the cache or didn't find it in cache
if not from_cache:
data = urlopen(url, timeout=3).read()
if cache is not None: # then put it in the cache
_cache_it(data, cache_path)
return data, from_cache
def _urlopen_cached(url, cache):
"""
Tries to load data from cache location otherwise downloads it. If it
downloads the data and cache is not None then it will put the downloaded
data in the cache path.
"""
from_cache = False
if cache is not None:
cache_path = join(cache, url.split("://")[-1].replace("/", ",") + ".zip")
try:
data = _open_cache(cache_path)
from_cache = True
except:
pass
# not using the cache or didn't find it in cache
if not from_cache:
data = urlopen(url).read()
if cache is not None: # then put it in the cache
_cache_it(data, cache_path)
return data, from_cache
def _urlopen_cached(url, cache):
"""
Tries to load data from cache location otherwise downloads it. If it
downloads the data and cache is not None then it will put the downloaded
data in the cache path.
"""
from_cache = False
if cache is not None:
cache_path = join(cache,
url.split("://")[-1].replace('/', ',') + ".zip")
try:
data = _open_cache(cache_path)
from_cache = True
except:
pass
# not using the cache or didn't find it in cache
if not from_cache:
data = urlopen(url).read()
if cache is not None: # then put it in the cache
_cache_it(data, cache_path)
return data, from_cache
else:
y = []
for _x in x:
y.append(fn(_x, **keywords))
y = np.array(y)
a = np.argsort(y)
return interp1d(y[a], x[a])
if __name__ == "__main__":
#TODO: Make sure everything is correctly aligned and make a plotting
# function
from statsmodels.compat.python import urlopen
import matplotlib.pyplot as plt
nerve_data = urlopen('http://www.statsci.org/data/general/nerve.txt')
nerve_data = np.loadtxt(nerve_data)
x = nerve_data / 50. # was in 1/50 seconds
cdf = ECDF(x)
x.sort()
F = cdf(x)
plt.step(x, F, where='post')
lower, upper = _conf_set(F)
plt.step(x, lower, 'r', where='post')
plt.step(x, upper, 'r', where='post')
plt.xlim(0, 1.5)
plt.ylim(0, 1.05)
plt.vlines(x, 0, .05)
plt.show()
y = fn(x, **keywords)
else:
y = []
for _x in x:
y.append(fn(_x, **keywords))
y = np.array(y)
a = np.argsort(y)
return interp1d(y[a], x[a])
if __name__ == "__main__":
#TODO: Make sure everything is correctly aligned and make a plotting
# function
from statsmodels.compat.python import urlopen
import matplotlib.pyplot as plt
nerve_data = urlopen('http://www.statsci.org/data/general/nerve.txt')
nerve_data = np.loadtxt(nerve_data)
x = nerve_data / 50. # was in 1/50 seconds
cdf = ECDF(x)
x.sort()
F = cdf(x)
plt.step(x, F, where='post')
lower, upper = _conf_set(F)
plt.step(x, lower, 'r', where='post')
plt.step(x, upper, 'r', where='post')
plt.xlim(0, 1.5)
plt.ylim(0, 1.05)
plt.vlines(x, 0, .05)
plt.show()
dt = (A + B) * 1 / thet
dq = (A + B) * np.log(lamb)
db = X * (dl * lamb)[:, np.newaxis]
sc = np.array([dt.sum(), dq.sum()])
sc = np.concatenate([db.sum(axis=0), sc])
return sc
#### Tests ####
from statsmodels.compat.python import urlopen
from numpy.testing import assert_almost_equal
import pandas
import patsy
medpar = pandas.read_csv(
urlopen(
'https://raw.githubusercontent.com/vincentarelbundock/Rdatasets/csv/COUNT/medpar.csv'
))
mdvis = pandas.read_csv(
urlopen(
'https://raw.githubusercontent.com/vincentarelbundock/Rdatasets/csv/COUNT/mdvis.csv'
))
# NB-2
'''
# R v2.15.1
library(MASS)
library(COUNT)
data(medpar)
f <- los~factor(type)+hmo+white
mod <- glm.nb(f, medpar)
summary(mod)
dl = (A + B) * Q / lamb - B * 1 / lamb
dt = (A + B) * 1 / thet
dq = (A + B) * np.log(lamb)
db = X * (dl * lamb)[:, np.newaxis]
sc = np.array([dt.sum(), dq.sum()])
sc = np.concatenate([db.sum(axis=0), sc])
return sc
#### Tests ####
from statsmodels.compat.python import urlopen
from numpy.testing import assert_almost_equal
import pandas
import patsy
medpar = pandas.read_csv(
urlopen(
'http://vincentarelbundock.github.com/Rdatasets/csv/COUNT/medpar.csv'))
mdvis = pandas.read_csv(
urlopen(
'http://vincentarelbundock.github.com/Rdatasets/csv/COUNT/mdvis.csv'))
# NB-2
'''
# R v2.15.1
library(MASS)
library(COUNT)
data(medpar)
f <- los~factor(type)+hmo+white
mod <- glm.nb(f, medpar)
summary(mod)
Call:
glm.nb(formula = f, data = medpar, init.theta = 2.243376203,
A = digamma(y+g) - digamma(g) + np.log(w)
B = g*(1-w) - y*w
dl = (A+B) * Q/lamb - B * 1/lamb
dt = (A+B) * 1/thet
dq = (A+B) * np.log(lamb)
db = X * (dl * lamb)[:,np.newaxis]
sc = np.array([dt.sum(), dq.sum()])
sc = np.concatenate([db.sum(axis=0), sc])
return sc
#### Tests ####
from statsmodels.compat.python import urlopen
from numpy.testing import assert_almost_equal
import pandas
import patsy
medpar = pandas.read_csv(urlopen('https://raw.githubusercontent.com/vincentarelbundock/Rdatasets/csv/COUNT/medpar.csv'))
mdvis = pandas.read_csv(urlopen('https://raw.githubusercontent.com/vincentarelbundock/Rdatasets/csv/COUNT/mdvis.csv'))
# NB-2
'''
# R v2.15.1
library(MASS)
library(COUNT)
data(medpar)
f <- los~factor(type)+hmo+white
mod <- glm.nb(f, medpar)
summary(mod)
Call:
glm.nb(formula = f, data = medpar, init.theta = 2.243376203,
link = log)
A = digamma(y+g) - digamma(g) + np.log(w)
B = g*(1-w) - y*w
dl = (A+B) * Q/lamb - B * 1/lamb
dt = (A+B) * 1/thet
dq = (A+B) * np.log(lamb)
db = X * (dl * lamb)[:,np.newaxis]
sc = np.array([dt.sum(), dq.sum()])
sc = np.concatenate([db.sum(axis=0), sc])
return sc
#### Tests ####
from statsmodels.compat.python import urlopen
from numpy.testing import assert_almost_equal
import pandas
import patsy
medpar = pandas.read_csv(urlopen('http://vincentarelbundock.github.com/Rdatasets/csv/COUNT/medpar.csv'))
mdvis = pandas.read_csv(urlopen('http://vincentarelbundock.github.com/Rdatasets/csv/COUNT/mdvis.csv'))
# NB-2
'''
# R v2.15.1
library(MASS)
library(COUNT)
data(medpar)
f <- los~factor(type)+hmo+white
mod <- glm.nb(f, medpar)
summary(mod)
Call:
glm.nb(formula = f, data = medpar, init.theta = 2.243376203,
link = log)
A = digamma(y+g) - digamma(g) + np.log(w)
B = g*(1-w) - y*w
dl = (A+B) * Q/lamb - B * 1/lamb
dt = (A+B) * 1/thet
dq = (A+B) * np.log(lamb)
db = X * (dl * lamb)[:,np.newaxis]
sc = np.array([dt.sum(), dq.sum()])
sc = np.concatenate([db.sum(axis=0), sc])
return sc
#### Tests ####
from statsmodels.compat.python import urlopen
from numpy.testing import assert_almost_equal
import pandas
import patsy
medpar = pandas.read_csv(urlopen('http://vincentarelbundock.github.com/Rdatasets/csv/COUNT/medpar.csv'))
mdvis = pandas.read_csv(urlopen('http://vincentarelbundock.github.com/Rdatasets/csv/COUNT/mdvis.csv'))
# NB-2
'''
# R v2.15.1
library(MASS)
library(COUNT)
data(medpar)
f <- los~factor(type)+hmo+white
mod <- glm.nb(f, medpar)
summary(mod)
Call:
glm.nb(formula = f, data = medpar, init.theta = 2.243376203,
link = log)
