def test_clone_duplicate_(self):
ds.df = self.df
ds2 = ds.clone_()
assert_frame_equal(ds2.df, ds.df)
ds2 = Ds()
# error clone
ds.df = None
ds2 = ds.clone_()
self.assertRaises(AttributeError)
ds.df = "wrong"
ds2 = ds.clone_()
self.assertRaises(AttributeError)
ds2 = ds._duplicate_(self.df, "db", False)
def test_concat(self):
df1 = pd.DataFrame([1, 2])
df2 = pd.DataFrame([3, 4])
ds.df = df1
ds2 = Ds(df2)
ds.concat(ds, ds2)
df3 = pd.DataFrame([1, 2, 3, 4])
self.assertEqual(list(ds.df), list(df3))
ds.df = df1
ds2 = ds.concat_(ds, ds2)
self.assertEqual(list(ds2.df), list(df3))
ds2 = None
self.assertErr(None, ds.concat_, ds, ds2)
self.assertErr(None, ds.concat, ds, ds2)
# -*- coding: utf-8 -*-
import pandas as pd
from pandas.testing import assert_frame_equal
from dataswim.tests.base import BaseDsTest
from dataswim import Ds
ds = Ds()
ds.errs_traceback = False
class TestDsDataText(BaseDsTest):
def test_flat(self):
df1 = pd.DataFrame([["one", "one"], ["two", "two"]],
columns=["one", "two"])
ds.df = df1
data = ds.flat_("one")
self.assertEqual(data, '0 one 1 two')
data = ds.flat_("one", False)
self.assertEqual(data, 'one two')
ds.df = None
data = ds.flat_("one")
self.assertRaises(TypeError)
"""def test_mfw(self):
df1 = pd.DataFrame([["one", "one"], ["two", "two"], ["one", "three"]],
columns=["one", "two"])
ds.df = df1
ds2 = ds.mfw_("one")
df2 = pd.DataFrame([[2, "one"], [1, "two"]],
columns=["Frequency", "Word"]).set_index('Word')
assert_frame_equal(ds2.df, df2)
def run(dbpath, debug=True):
ds = Ds()
if debug is True:
ds.status("Debug mode is enabled")
db.init(dbpath)
results, modified_repos = get_results()
if debug is False:
ds.quiet = True
modulepath = os.path.dirname(os.path.realpath(ghobserver.__file__))
# templates_path = modulepath + "/templates/charts"
static_path = modulepath + "/static"
repos = db.get_repos()
ds.connect("sqlite:///" + dbpath)
ds.load("gh_commit")
ds.relation("repository", "repository", "name", "Repository")
ds.rename("date", "Date")
ds.rename("additions", "Additions")
ds.rename("deletions", "Deletions")
ds.rename("changed_files", "Changed files")
ds.keep("Repository", "Date", "Additions", "Deletions", "Changed files")
ds.date("Date")
ds.dateindex("Date")
ds.report_path = static_path + "/charts"
ds.static_path = ds.report_path
ds.backup()
ts = [["3Y", "1M"], ["1Y", "1W"], ["3M", "1D"], ["3W", "1D"], ["1W", "1D"],
["100Y", "1M"]]
# repos = [repos[0]]
for tf in ts:
timeframe = tf[0]
timerange = tf[1]
ds.restore()
# all repos
slug = "all"
# ds.append(["", now(), 0, 0, 0])
ds.df.Date = pd.to_datetime(ds.df.Date, utc=True)
ds.dateindex("Date")
ds.nowrange("Date", timeframe)
if len(ds.df.index) < 2:
no_data(ds, slug, timeframe)
ds.status("No data for all repositories", timeframe)
else:
# add repositories names to the rsumed dataset
res = []
dss = ds.split_("Repository")
for k in dss:
d = dss[k]
repo = d.df["Repository"].values[0]
try:
d.rsum(timerange, "Commits")
except Exception as e:
raise (e)
try:
d.add("Repository", repo)
except Exception as e:
raise (e)
res.append(d)
ds2 = Ds().concat_(*res)
ds.rsum(timerange, "Commits")
ds.df.iloc[[-1], [3]] = 0
ds.indexcol("Date")
pipe_repo(ds, None, timeframe, ds2)
# by repos
for repo in repos:
reposlug = repo["name"]
if debug is False:
if reposlug not in modified_repos:
continue
ds.restore()
try:
ds.append([reposlug, now(), 0, 0, 0])
except Exception as e:
raise (e)
ds.df.Date = pd.to_datetime(ds.df.Date)
ds.dateindex("Date")
ds.exact("Repository", reposlug)
ds.nowrange("Date", timeframe)
if len(ds.df.index) < 2:
no_data(ds, reposlug, timeframe)
ds.status("No data for", reposlug, timeframe)
continue
ds.rsum(timerange, "Commits")
ds.indexcol("Date")
ds.df.iloc[[-1], [3]] = 0
ds, slug = pipe_repo(ds, repo["name"], timeframe)
if results:
pop_notification(results)
db.clean_results()
print("ok")