Skip to content

int - np.int64 #158

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
jwijffels wants to merge 1 commit into
base: main
Choose a base branch
from
Open

int - np.int64 #158

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
86 changes: 43 additions & 43 deletions adelie/state.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -50,7 +50,7 @@ def render_dual_groups(
):
return np.cumsum(np.concatenate(
[[0] + [0 if c is None else c.dual_size for c in constraints]],
dtype=int,
dtype=np.int64,
))[:-1]


Expand All @@ -68,7 +68,7 @@ def deduce_states(
)
screen_begins = np.cumsum(
np.concatenate([[0], group_sizes[screen_set]]),
dtype=int,
dtype=np.int64,
)[:-1]
return (
constraints,
Expand Down Expand Up @@ -256,7 +256,7 @@ def check(

# ================ screen_begins check ====================
expected = np.cumsum(
np.concatenate([[0], self.group_sizes[self.screen_set]], dtype=int)
np.concatenate([[0], self.group_sizes[self.screen_set]], dtype=np.int64)
)
WS = expected[-1]
expected = expected[:-1]
Expand Down Expand Up @@ -333,7 +333,7 @@ def check(

# ================ active_begins check ====================
expected = np.cumsum(
np.concatenate([[0], self.group_sizes[self.screen_set[active_set]]], dtype=int)
np.concatenate([[0], self.group_sizes[self.screen_set[active_set]]], dtype=np.int64)
)
WA = expected[-1]
expected = expected[:-1]
Expand Down Expand Up @@ -580,7 +580,7 @@ def gaussian_pin_naive(
)

p = X.cols()
group_sizes = np.concatenate([groups, [p]], dtype=int)
group_sizes = np.concatenate([groups, [p]], dtype=np.int64)
group_sizes = group_sizes[1:] - group_sizes[:-1]

dtype = (
Expand All @@ -603,17 +603,17 @@ def __init__(self):
# static inputs require a reference to input
# or copy if it must be made
self._X = X
self._groups = np.array(groups, copy=True, dtype=int)
self._group_sizes = np.array(group_sizes, copy=True, dtype=int)
self._groups = np.array(groups, copy=True, dtype=np.int64)
self._group_sizes = np.array(group_sizes, copy=True, dtype=np.int64)
self._penalty = np.array(penalty, copy=True, dtype=dtype)
self._weights = np.array(weights, copy=True, dtype=dtype)
self._screen_set = np.array(screen_set, copy=True, dtype=int)
self._screen_set = np.array(screen_set, copy=True, dtype=np.int64)
self._lmda_path = np.array(lmda_path, copy=True, dtype=dtype)
# dynamic inputs require a copy to not modify user's inputs
self._resid = np.array(resid, copy=True, dtype=dtype)
self._screen_beta = np.array(screen_beta, copy=True, dtype=dtype)
self._screen_is_active = np.array(screen_is_active, copy=True, dtype=bool)
self._active_set = np.array(active_set, copy=True, dtype=int)
self._active_set = np.array(active_set, copy=True, dtype=np.int64)

(
self._constraints,
Expand Down Expand Up @@ -879,7 +879,7 @@ def gaussian_pin_cov(
)

p = A.cols()
group_sizes = np.concatenate([groups, [p]], dtype=int)
group_sizes = np.concatenate([groups, [p]], dtype=np.int64)
group_sizes = group_sizes[1:] - group_sizes[:-1]

dtype = (
Expand All @@ -902,16 +902,16 @@ def __init__(self):
# static inputs require a reference to input
# or copy if it must be made
self._A = A
self._groups = np.array(groups, copy=True, dtype=int)
self._group_sizes = np.array(group_sizes, copy=True, dtype=int)
self._groups = np.array(groups, copy=True, dtype=np.int64)
self._group_sizes = np.array(group_sizes, copy=True, dtype=np.int64)
self._penalty = np.array(penalty, copy=True, dtype=dtype)
self._screen_set = np.array(screen_set, copy=True, dtype=int)
self._screen_set = np.array(screen_set, copy=True, dtype=np.int64)
self._lmda_path = np.array(lmda_path, copy=True, dtype=dtype)
# dynamic inputs require a copy to not modify user's inputs
self._screen_beta = np.array(screen_beta, copy=True, dtype=dtype)
self._screen_grad = np.array(screen_grad, copy=True, dtype=dtype)
self._screen_is_active = np.array(screen_is_active, copy=True, dtype=bool)
self._active_set = np.array(active_set, copy=True, dtype=int)
self._active_set = np.array(active_set, copy=True, dtype=np.int64)

(
self._constraints,
Expand Down Expand Up @@ -954,7 +954,7 @@ def __init__(self):
self._screen_subset = np.concatenate([
np.arange(groups[i], groups[i] + group_sizes[i])
for i in self._screen_set
], dtype=int)
], dtype=np.int64)
self._screen_subset_order = np.argsort(self._screen_subset)
self._screen_subset_ordered = self._screen_subset[self._screen_subset_order]

Expand Down Expand Up @@ -1352,16 +1352,16 @@ def __init__(self):
self._A = A
self._v = np.array(v, copy=True, dtype=dtype)
self._constraints = render_constraints(groups.shape[0], constraints, dtype)
self._groups = np.array(groups, copy=True, dtype=int)
self._group_sizes = np.array(group_sizes, copy=True, dtype=int)
self._groups = np.array(groups, copy=True, dtype=np.int64)
self._group_sizes = np.array(group_sizes, copy=True, dtype=np.int64)
self._dual_groups = render_dual_groups(self._constraints)
self._penalty = np.array(penalty, copy=True, dtype=dtype)
self._lmda_path = np.array(lmda_path, copy=False, dtype=dtype)
self._screen_set = np.array(screen_set, copy=False, dtype=int)
self._screen_set = np.array(screen_set, copy=False, dtype=np.int64)
self._screen_beta = np.array(screen_beta, copy=False, dtype=dtype)
self._screen_is_active = np.array(screen_is_active, copy=False, dtype=bool)
self._grad = np.array(grad, copy=False, dtype=dtype)
self._active_set = np.array(active_set, copy=False, dtype=int)
self._active_set = np.array(active_set, copy=False, dtype=np.int64)

# MUST call constructor directly and not use super()!
# https://pybind11.readthedocs.io/en/stable/advanced/classes.html#forced-trampoline-class-initialisation
Expand Down Expand Up @@ -1519,7 +1519,7 @@ def check(

# ================ screen_begins check ====================
expected = np.cumsum(
np.concatenate([[0], self.group_sizes[self.screen_set]], dtype=int)
np.concatenate([[0], self.group_sizes[self.screen_set]], dtype=np.int64)
)
WS = expected[-1]
expected = expected[:-1]
Expand Down Expand Up @@ -1553,7 +1553,7 @@ def check(
self.group_sizes[self.screen_set],
)
if np.any(self.screen_beta[sb:sb+gs] != 0)
], dtype=int)
], dtype=np.int64)
self._check(
np.all(self.screen_is_active[nnz_idxs]),
"check screen_is_active is only active on non-zeros of screen_beta",
Expand All @@ -1572,9 +1572,9 @@ def check(
self.X.btmul(g, gs, self.screen_beta[b:b+gs], Xbeta)

if len(screen_indices) == 0:
screen_indices = np.array(screen_indices, dtype=int)
screen_indices = np.array(screen_indices, dtype=np.int64)
else:
screen_indices = np.concatenate(screen_indices, dtype=int)
screen_indices = np.concatenate(screen_indices, dtype=np.int64)

resid = yc - Xbeta
grad = np.empty(p)
Expand Down Expand Up @@ -1948,16 +1948,16 @@ def __init__(self):
self._X = X
self._X_means = np.array(X_means, copy=True, dtype=dtype)
self._constraints = render_constraints(groups.shape[0], constraints, dtype)
self._groups = np.array(groups, copy=True, dtype=int)
self._group_sizes = np.array(group_sizes, copy=True, dtype=int)
self._groups = np.array(groups, copy=True, dtype=np.int64)
self._group_sizes = np.array(group_sizes, copy=True, dtype=np.int64)
self._dual_groups = render_dual_groups(self._constraints)
self._penalty = np.array(penalty, copy=True, dtype=dtype)
self._offsets = np.array(offsets, copy=True, dtype=dtype)
self._lmda_path = np.array(lmda_path, copy=False, dtype=dtype)
self._screen_set = np.array(screen_set, copy=False, dtype=int)
self._screen_set = np.array(screen_set, copy=False, dtype=np.int64)
self._screen_beta = np.array(screen_beta, copy=False, dtype=dtype)
self._screen_is_active = np.array(screen_is_active, copy=False, dtype=bool)
self._active_set = np.array(active_set, copy=False, dtype=int)
self._active_set = np.array(active_set, copy=False, dtype=np.int64)
self._grad = np.array(grad, copy=False, dtype=dtype)
self._resid = np.array(resid, copy=False, dtype=dtype)

Expand Down Expand Up @@ -2308,17 +2308,17 @@ def __init__(self):
self._X_expanded = X
self._X_means = np.array(X_means, copy=True, dtype=dtype)
self._constraints = render_constraints(groups.shape[0], constraints, dtype)
self._groups = np.array(groups, copy=True, dtype=int)
self._group_sizes = np.array(group_sizes, copy=True, dtype=int)
self._groups = np.array(groups, copy=True, dtype=np.int64)
self._group_sizes = np.array(group_sizes, copy=True, dtype=np.int64)
self._dual_groups = render_dual_groups(self._constraints)
self._penalty = np.array(penalty, copy=True, dtype=dtype)
self._weights_expanded = np.repeat(self._glm.weights, repeats=n_classes) / n_classes
self._offsets = np.array(offsets, copy=True, dtype=dtype)
self._lmda_path = np.array(lmda_path, copy=False, dtype=dtype)
self._screen_set = np.array(screen_set, copy=False, dtype=int)
self._screen_set = np.array(screen_set, copy=False, dtype=np.int64)
self._screen_beta = np.array(screen_beta, copy=False, dtype=dtype)
self._screen_is_active = np.array(screen_is_active, copy=False, dtype=bool)
self._active_set = np.array(active_set, copy=False, dtype=int)
self._active_set = np.array(active_set, copy=False, dtype=np.int64)
self._grad = np.array(grad, copy=False, dtype=dtype)
self._resid = np.array(resid, copy=False, dtype=dtype)

Expand Down Expand Up @@ -2675,16 +2675,16 @@ def __init__(self):
self._glm = glm
self._X = X
self._constraints = render_constraints(groups.shape[0], constraints, dtype)
self._groups = np.array(groups, copy=True, dtype=int)
self._group_sizes = np.array(group_sizes, copy=True, dtype=int)
self._groups = np.array(groups, copy=True, dtype=np.int64)
self._group_sizes = np.array(group_sizes, copy=True, dtype=np.int64)
self._dual_groups = render_dual_groups(self._constraints)
self._penalty = np.array(penalty, copy=True, dtype=dtype)
self._offsets = np.array(offsets, copy=True, dtype=dtype)
self._lmda_path = np.array(lmda_path, copy=False, dtype=dtype)
self._screen_set = np.array(screen_set, copy=False, dtype=int)
self._screen_set = np.array(screen_set, copy=False, dtype=np.int64)
self._screen_beta = np.array(screen_beta, copy=False, dtype=dtype)
self._screen_is_active = np.array(screen_is_active, copy=False, dtype=bool)
self._active_set = np.array(active_set, copy=False, dtype=int)
self._active_set = np.array(active_set, copy=False, dtype=np.int64)
self._grad = np.array(grad, copy=False, dtype=dtype)
self._eta = np.array(eta, copy=False, dtype=dtype)
self._resid = np.array(resid, copy=False, dtype=dtype)
Expand Down Expand Up @@ -3041,16 +3041,16 @@ def __init__(self):
self._X = X_raw
self._X_expanded = X
self._constraints = render_constraints(groups.shape[0], constraints, dtype)
self._groups = np.array(groups, copy=True, dtype=int)
self._group_sizes = np.array(group_sizes, copy=True, dtype=int)
self._groups = np.array(groups, copy=True, dtype=np.int64)
self._group_sizes = np.array(group_sizes, copy=True, dtype=np.int64)
self._dual_groups = render_dual_groups(self._constraints)
self._penalty = np.array(penalty, copy=True, dtype=dtype)
self._offsets = np.array(offsets, copy=True, dtype=dtype)
self._lmda_path = np.array(lmda_path, copy=False, dtype=dtype)
self._screen_set = np.array(screen_set, copy=False, dtype=int)
self._screen_set = np.array(screen_set, copy=False, dtype=np.int64)
self._screen_beta = np.array(screen_beta, copy=False, dtype=dtype)
self._screen_is_active = np.array(screen_is_active, copy=False, dtype=bool)
self._active_set = np.array(active_set, copy=False, dtype=int)
self._active_set = np.array(active_set, copy=False, dtype=np.int64)
self._grad = np.array(grad, copy=False, dtype=dtype)
self._eta = np.array(eta, copy=False, dtype=dtype)
self._resid = np.array(resid, copy=False, dtype=dtype)
Expand Down Expand Up @@ -3228,9 +3228,9 @@ def __init__(self):
self._lower = np.array(lower, copy=False, dtype=dtype)
self._upper = np.array(upper, copy=False, dtype=dtype)
self._weights = np.array(weights, copy=False, dtype=dtype)
self._screen_set = np.array(screen_set, copy=True, dtype=int)
self._screen_set = np.array(screen_set, copy=True, dtype=np.int64)
self._is_screen = np.array(is_screen, copy=True, dtype=bool)
self._active_set = np.array(active_set, copy=True, dtype=int)
self._active_set = np.array(active_set, copy=True, dtype=np.int64)
self._is_active = np.array(is_active, copy=True, dtype=bool)
self._beta = np.array(beta, copy=True, dtype=dtype)
self._resid = np.array(resid, copy=True, dtype=dtype)
Expand Down Expand Up @@ -3383,11 +3383,11 @@ def __init__(self):
self._S = np.array(S, copy=False, dtype=dtype, order="F")
self._penalty_neg = np.array(penalty_neg, copy=False, dtype=dtype)
self._penalty_pos = np.array(penalty_pos, copy=False, dtype=dtype)
self._screen_set = np.array(screen_set, copy=True, dtype=int)
self._screen_set = np.array(screen_set, copy=True, dtype=np.int64)
self._is_screen = np.array(is_screen, copy=True, dtype=bool)
self._screen_ASAT_diag = np.array(screen_ASAT_diag, copy=True, dtype=dtype)
self._screen_AS = np.array(screen_AS, copy=True, dtype=dtype, order="C")
self._active_set = np.array(active_set, copy=True, dtype=int)
self._active_set = np.array(active_set, copy=True, dtype=np.int64)
self._is_active = np.array(is_active, copy=True, dtype=bool)
self._beta = np.array(beta, copy=True, dtype=dtype)
self._resid = np.array(resid, copy=True, dtype=dtype)
Expand Down