def test_std():
(chunk, chunk_expr), (agg, agg_expr) = split(t, t.amount.std())
assert chunk.schema == t.schema
assert chunk_expr.isidentical(summary(x=chunk.amount.sum(),
x2=(chunk.amount ** 2).sum(),
n=chunk.amount.count(),
keepdims=True))
assert isrecord(agg.dshape.measure)
assert agg_expr.isidentical(sqrt((agg.x2.sum() / (agg.n.sum())
- (agg.x.sum() / (agg.n.sum())) ** 2)))
def test_std():
(chunk, chunk_expr), (agg, agg_expr) = split(t, t.amount.std())
assert chunk.schema == t.schema
assert chunk_expr.isidentical(
summary(x=chunk.amount.sum(),
x2=(chunk.amount**2).sum(),
n=chunk.amount.count(),
keepdims=True))
assert isrecord(agg.dshape.measure)
assert agg_expr.isidentical(
sqrt(agg.x2.sum() / agg.n.sum() - (agg.x.sum() / (agg.n.sum()))**2))
def test_complex_summaries():
t = symbol('t', '100 * {a: int, b: int}')
(chunk, chunk_expr), (agg, agg_expr) = split(
t, summary(q=t.a.mean(), w=t.a.std(), e=t.a.sum()))
assert chunk_expr.isidentical(
summary(e=chunk.a.sum(),
q_count=chunk.a.count(),
q_total=chunk.a.sum(),
w_n=chunk.a.count(),
w_x=chunk.a.sum(),
w_x2=(chunk.a**2).sum(),
keepdims=True))
expected = summary(e=agg.e.sum(),
q=agg.q_total.sum() / agg.q_count.sum(),
w=sqrt((agg.w_x2.sum() / agg.w_n.sum()) -
(agg.w_x.sum() / agg.w_n.sum())**2))
assert agg_expr.isidentical(expected)
def test_complex_summaries():
t = symbol('t', '100 * {a: int, b: int}')
(chunk, chunk_expr), (agg, agg_expr) = split(t, summary(q=t.a.mean(),
w=t.a.std(),
e=t.a.sum()))
assert chunk_expr.isidentical(summary(e=chunk.a.sum(),
q_count=chunk.a.count(),
q_total=chunk.a.sum(),
w_n=chunk.a.count(),
w_x=chunk.a.sum(),
w_x2=(chunk.a ** 2).sum(),
keepdims=True))
expected = summary(e=agg.e.sum(),
q=agg.q_total.sum() / agg.q_count.sum(),
w=sqrt((agg.w_x2.sum() / agg.w_n.sum())
- (agg.w_x.sum() / agg.w_n.sum()) ** 2))
assert agg_expr.isidentical(expected)