1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
|
// Copyright 2018 Global Phasing Ltd.
#include <complex>
// for symmetrize_min and symmetrize_max
bool operator<(const std::complex<float>& a, const std::complex<float>& b) {
return std::norm(a) < std::norm(b);
}
bool operator>(const std::complex<float>& a, const std::complex<float>& b) {
return std::norm(a) > std::norm(b);
}
#include "common.h"
#include "array.h"
#include "make_iterator.h"
#include <nanobind/ndarray.h>
#include <nanobind/stl/array.h>
#include <nanobind/stl/complex.h>
#include <nanobind/stl/string.h>
#include <nanobind/stl/vector.h> // for find_blobs_by_flood_fill, ...
#include "gemmi/grid.hpp"
#include "gemmi/floodfill.hpp" // for flood_fill_above
#include "gemmi/solmask.hpp" // for SolventMasker, mask_points_in_constant_radius
#include "gemmi/blob.hpp" // for Blob, find_blobs_by_flood_fill
#include "gemmi/asumask.hpp" // for MaskedGrid
using namespace gemmi;
namespace {
template<typename T>
auto grid_to_array(GridMeta& g, std::vector<T>& data) {
// should we take AxisOrder into account here?
return nb::ndarray<nb::numpy, T>(data.data(),
{(size_t)g.nu, (size_t)g.nv, (size_t)g.nw},
nb::handle(),
{1, g.nu, g.nu * g.nv});
}
template<typename T>
nb::class_<GridBase<T>, GridMeta> add_grid_base(nb::module_& m, const char* name) {
using GrBase = GridBase<T>;
using GrPoint = typename GridBase<T>::Point;
nb::class_<GrBase, GridMeta> grid_base(m, name);
nb::class_<GrPoint>(grid_base, "Point")
.def_ro("u", &GrPoint::u)
.def_ro("v", &GrPoint::v)
.def_ro("w", &GrPoint::w)
.def_prop_rw("value",
[](const GrPoint& self) { return *self.value; },
[](GrPoint& self, T x) { *self.value = x; })
.def("__repr__", [=](const GrPoint& self) {
return nb::str("<gemmi.{}.Point ({}, {}, {}) -> {}>")
.format(name, self.u, self.v, self.w, self.value);
});
auto to_array = [](GrBase& gr) { return grid_to_array(gr, gr.data); };
grid_base
.def_prop_ro("array", to_array, nb::rv_policy::reference_internal)
.def("__array__", [](nb::handle_t<GrBase>& h, nb::handle dtype, nb::handle copy) {
return handle_numpy_array_args(h.attr("array"), dtype, copy);
}, nb::arg("dtype")=nb::none(), nb::arg("copy")=nb::none())
.def("point_to_index", &GrBase::point_to_index)
.def("index_to_point", &GrBase::index_to_point)
.def("fill", &GrBase::fill, nb::arg("value"))
.def("sum", &GrBase::sum)
.def("__iter__", [](GrBase& self) {
return usual_iterator(self, self);
}, nb::keep_alive<0, 1>())
;
return grid_base;
}
template<typename T>
nb::class_<Grid<T>, GridBase<T>> add_grid_common(nb::module_& m, const std::string& name) {
using Gr = Grid<T>;
using GrPoint = typename GridBase<T>::Point;
using Masked = MaskedGrid<T>;
nb::class_<Gr, GridBase<T>> grid(m, name.c_str());
nb::class_<Masked> masked_grid (m, ("Masked" + name).c_str());
grid
.def(nb::init<>())
.def("__init__", [](Gr* grid, int nx, int ny, int nz) {
new(grid) Gr();
grid->set_size(nx, ny, nz);
}, nb::arg("nx"), nb::arg("ny"), nb::arg("nz"))
.def("__init__", [](Gr* grid, const nb::ndarray<nb::numpy, T, nb::ndim<3>>& arr,
const UnitCell *cell, const SpaceGroup* sg) {
new(grid) Gr();
auto r = arr.view();
grid->set_size((int)r.shape(0), (int)r.shape(1), (int)r.shape(2));
for (size_t k = 0; k < r.shape(2); ++k)
for (size_t j = 0; j < r.shape(1); ++j)
for (size_t i = 0; i < r.shape(0); ++i)
grid->data[grid->index_q(i, j, k)] = r(i, j, k);
if (cell)
grid->set_unit_cell(*cell);
if (sg)
grid->spacegroup = sg;
}, nb::arg().noconvert(), nb::arg("cell")=nb::none(), nb::arg("spacegroup")=nb::none())
.def_prop_ro("spacing", [](const Gr& self) {
return nb::make_tuple(self.spacing[0], self.spacing[1], self.spacing[2]);
})
.def("set_size", &Gr::set_size)
.def("set_size_from_spacing", &Gr::set_size_from_spacing,
nb::arg("spacing"), nb::arg("rounding"))
.def("get_value", &Gr::get_value)
.def("set_value", &Gr::set_value)
.def("get_point", &Gr::get_point)
.def("get_nearest_point", (GrPoint (Gr::*)(const Position&)) &Gr::get_nearest_point)
.def("point_to_fractional", &Gr::point_to_fractional)
.def("point_to_position", &Gr::point_to_position)
.def("change_values", &Gr::change_values, nb::arg("old_value"), nb::arg("new_value"))
.def("copy_metadata_from", &Gr::copy_metadata_from)
.def("setup_from", &Gr::template setup_from<Structure>,
nb::arg("st"), nb::arg("spacing")=0.)
.def("set_unit_cell", (void (Gr::*)(const UnitCell&)) &Gr::set_unit_cell)
.def("set_points_around", &Gr::set_points_around,
nb::arg("position"), nb::arg("radius"), nb::arg("value"), nb::arg("use_pbc")=true)
.def("symmetrize_min", &Gr::symmetrize_min)
.def("symmetrize_max", &Gr::symmetrize_max)
.def("symmetrize_abs_max", &Gr::symmetrize_abs_max)
.def("symmetrize_sum", &Gr::symmetrize_sum)
.def("masked_asu", &masked_asu<T>, nb::keep_alive<0, 1>())
.def("mask_points_in_constant_radius", &mask_points_in_constant_radius<T>,
nb::arg("model"), nb::arg("radius"), nb::arg("value"),
nb::arg("ignore_hydrogen")=false, nb::arg("ignore_zero_occupancy_atoms")=false)
.def("get_subarray",
[](const Gr& self, std::array<int,3> start, std::array<int,3> shape) {
auto arr = make_numpy_array<T>(
{(size_t)shape[0], (size_t)shape[1], (size_t)shape[2]},
{1, int64_t(shape[0]), int64_t(shape[0]*shape[1])});
self.get_subarray(arr.data(), start, shape);
return arr;
}, nb::arg("start"), nb::arg("shape"))
.def("set_subarray",
[](Gr& self,
const nb::ndarray<T, nb::ndim<3>, nb::f_contig, nb::device::cpu>& arr,
std::array<int,3> start) {
self.set_subarray(arr.data(), start,
{(int)arr.shape(0), (int)arr.shape(1), (int)arr.shape(2)});
}, nb::arg("arr"), nb::arg("start"))
.def("clone", [](const Gr& self) { return new Gr(self); })
.def("__repr__", [=](const Gr& self) {
return cat("<gemmi.", name, '(', self.nu, ", ", self.nv, ", ", self.nw, ")>");
});
masked_grid
.def_ro("grid", &Masked::grid, nb::rv_policy::reference)
.def_prop_ro("mask_array", [](Masked& self) {
return grid_to_array(*self.grid, self.mask);
}, nb::rv_policy::reference_internal)
.def("__iter__", [](Masked& self) {
return usual_iterator(self, self);
}, nb::keep_alive<0, 1>())
;
return grid;
}
template<typename T>
void add_grid_interpolation(nb::class_<Grid<T>, GridBase<T>>& grid) {
using Gr = Grid<T>;
grid
.def("interpolate_value",
(T (Gr::*)(const Fractional&, int) const) &Gr::interpolate_value,
nb::arg(), nb::arg("order")=1)
.def("interpolate_value",
(T (Gr::*)(const Position&, int) const) &Gr::interpolate_value,
nb::arg(), nb::arg("order")=1)
// deprecated, use interpolate_value(..., order=3)
.def("tricubic_interpolation",
(double (Gr::*)(const Fractional&) const) &Gr::tricubic_interpolation)
// deprecated, use interpolate_value(..., order=3)
.def("tricubic_interpolation",
(double (Gr::*)(const Position&) const) &Gr::tricubic_interpolation)
.def("tricubic_interpolation_der",
(std::array<double,4> (Gr::*)(const Fractional&) const)
&Gr::tricubic_interpolation_der)
.def("interpolate_position_array",
[](const Gr& self, const nb::ndarray<double, nb::shape<-1,3>, nb::device::cpu>& xyz,
int order, const Transform* to_frac) {
auto xyz_view = xyz.view();
size_t len = xyz_view.shape(0);
auto values = make_numpy_array<T>({len});
T* data = values.data();
const Transform& frac = to_frac ? *to_frac : self.unit_cell.frac;
for (size_t i = 0; i < len; ++i) {
Position pos(xyz_view(i, 0), xyz_view(i, 1), xyz_view(i, 2));
Fractional fpos = Fractional(frac.apply(pos));
data[i] = self.interpolate_value(fpos, order);
}
return values;
}, nb::arg("xyz"), nb::arg("order")=1, nb::arg("to_frac")=nb::none())
// The name of this function is not very descriptive, but since it's used
// in a few external projects, renaming it isn't worth the hassle.
// cf. interpolate_grid
.def("interpolate_values",
[](const Gr& self, nb::ndarray<nb::numpy, T, nb::ndim<3>>& arr,
const Transform& tr, int order) {
auto r = arr.view();
for (size_t i = 0; i < r.shape(0); ++i)
for (size_t j = 0; j < r.shape(1); ++j)
for (size_t k = 0; k < r.shape(2); ++k) {
Position pos(tr.apply(Vec3(i, j, k)));
Fractional fpos = self.unit_cell.fractionalize(pos);
r(i, j, k) = self.interpolate_value(fpos, order);
}
}, nb::arg().noconvert(), nb::arg(), nb::arg("order")=1)
.def("interpolate_grid_flexible",
// Interpolate the moving grid onto the interpolated grid, applying different transforms to
// different regions
[](
Gr& interpolated_map,
const Gr& moving_map,
std::vector<nb::ndarray<int, nb::shape<-1,3>>>& point_arr_vec,
std::vector<nb::ndarray<float, nb::shape<-1,3>>>& pos_arr_vec,
std::vector<Transform> transform_vec
) {
// For each transform in the list, interpolate the positions in the moving map
// corresponding to points in the interpolated map. This method is based on
// contiguous groups of positions in real space, and hence the need to
// separate out the point and pos arrays, to ensure the correct symmetry
// copies of of the positions corresponding to a map point are used.
// Python code to derive such partitionings of the unit cell around a
// molecule can be found in https://github.com/ConorFWild/pandda_2_gemmi
for (std::size_t i=0; i < point_arr_vec.size(); i++){
nb::ndarray<int, nb::shape<-1,3>> point_array = point_arr_vec[i];
nb::ndarray<float, nb::shape<-1,3>> pos_array = pos_arr_vec[i];
const Transform transform = transform_vec[i];
auto r_point = point_array.view();
auto r_pos = pos_array.view();
for (std::size_t i=0; i < r_point.shape(0); i++)
{
// Position
Position pos = Position(
r_pos(i,0),
r_pos(i,1),
r_pos(i,2)
);
//transform
Position pos_moving = Position(transform.apply(pos));
// fractionalise
Fractional pos_moving_fractional = moving_map.unit_cell.fractionalize(pos_moving);
// interpolate
float interpolated_value = moving_map.interpolate_value(pos_moving_fractional);
// assign
interpolated_map.set_value(
r_point(i, 0),
r_point(i, 1),
r_point(i, 2),
interpolated_value
);
};
}
}, nb::arg(), nb::arg().noconvert(), nb::arg().noconvert(), nb::arg())
;
}
} // anonymous namespace
void add_grid(nb::module_& m) {
nb::enum_<AxisOrder>(m, "AxisOrder")
.value("Unknown", AxisOrder::Unknown)
.value("XYZ", AxisOrder::XYZ)
.value("ZYX", AxisOrder::ZYX);
nb::enum_<GridSizeRounding>(m, "GridSizeRounding")
.value("Nearest", GridSizeRounding::Nearest)
.value("Up", GridSizeRounding::Up)
.value("Down", GridSizeRounding::Down);
nb::class_<GridMeta>(m, "GridMeta")
.def_rw("spacegroup", &GridMeta::spacegroup)
.def_rw("unit_cell", &GridMeta::unit_cell)
.def_ro("nu", &GridMeta::nu, "size in the first (fastest-changing) dim")
.def_ro("nv", &GridMeta::nv, "size in the second dimension")
.def_ro("nw", &GridMeta::nw, "size in the third (slowest-changing) dim")
.def_ro("axis_order", &GridMeta::axis_order)
.def_prop_ro("point_count", &GridMeta::point_count)
.def("get_position", &GridMeta::get_position)
.def("get_fractional", &GridMeta::get_fractional)
.def_prop_ro("shape", [](const GridMeta& self) {
return nb::make_tuple(self.nu, self.nv, self.nw);
});
add_grid_base<int8_t>(m, "Int8GridBase")
.def("get_nonzero_extent", &get_nonzero_extent<int8_t>);
add_grid_common<int8_t>(m, "Int8Grid");
add_grid_base<float>(m, "FloatGridBase")
.def("calculate_correlation", &calculate_correlation<float>)
.def("get_nonzero_extent", &get_nonzero_extent<float>)
;
auto grid_float = add_grid_common<float>(m, "FloatGrid");
add_grid_interpolation<float>(grid_float);
grid_float.def("symmetrize_avg", &Grid<float>::symmetrize_avg);
grid_float.def("normalize", &Grid<float>::normalize);
grid_float.def("add_soft_edge_to_mask", &add_soft_edge_to_mask<float>);
add_grid_base<std::complex<float>>(m, "ComplexGridBase");
// from solmask.hpp
nb::enum_<AtomicRadiiSet>(m, "AtomicRadiiSet")
.value("VanDerWaals", AtomicRadiiSet::VanDerWaals)
.value("Cctbx", AtomicRadiiSet::Cctbx)
.value("Refmac", AtomicRadiiSet::Refmac)
.value("Constant", AtomicRadiiSet::Constant);
nb::class_<SolventMasker>(m, "SolventMasker")
.def(nb::init<AtomicRadiiSet, double>(),
nb::arg("choice"), nb::arg("constant_r")=0.)
.def_rw("atomic_radii_set", &SolventMasker::atomic_radii_set)
.def_rw("rprobe", &SolventMasker::rprobe)
.def_rw("rshrink", &SolventMasker::rshrink)
.def_rw("island_min_volume", &SolventMasker::island_min_volume)
.def_rw("constant_r", &SolventMasker::constant_r)
.def_rw("ignore_hydrogen", &SolventMasker::ignore_hydrogen)
.def_rw("ignore_zero_occupancy_atoms", &SolventMasker::ignore_zero_occupancy_atoms)
.def_rw("use_atom_occupancy", &SolventMasker::use_atom_occupancy)
.def("set_radii", &SolventMasker::set_radii,
nb::arg("choice"), nb::arg("constant_r")=0.)
.def("put_mask_on_int8_grid", &SolventMasker::put_mask_on_grid<int8_t>)
.def("put_mask_on_float_grid", &SolventMasker::put_mask_on_grid<float>)
.def("set_to_zero", &SolventMasker::set_to_zero)
;
m.def("interpolate_grid", &interpolate_grid<float>,
nb::arg("dest"), nb::arg("src"), nb::arg("tr"), nb::arg("order")=1);
m.def("interpolate_grid_around_model", &interpolate_grid_around_model<float>,
nb::arg("dest"), nb::arg("src"), nb::arg("tr"),
nb::arg("dest_model"), nb::arg("radius"), nb::arg("order")=1);
// from blob.hpp
nb::class_<Blob>(m, "Blob")
.def_ro("volume", &Blob::volume)
.def_ro("score", &Blob::score)
.def_ro("peak_value", &Blob::peak_value)
.def_ro("centroid", &Blob::centroid)
.def_ro("peak_pos", &Blob::peak_pos)
;
m.def("find_blobs_by_flood_fill",
[](const Grid<float>& grid, double cutoff, double min_volume,
double min_score, double min_peak, bool negate) {
BlobCriteria crit;
crit.cutoff = cutoff;
crit.min_volume = min_volume;
crit.min_score = min_score;
crit.min_peak = min_peak;
return find_blobs_by_flood_fill(grid, crit, negate);
}, nb::arg("grid"), nb::arg("cutoff"), nb::arg("min_volume")=10.,
nb::arg("min_score")=15., nb::arg("min_peak")=0., nb::arg("negate")=false);
// from floodfill.hpp
m.def("flood_fill_above", &flood_fill_above,
nb::arg("grid"), nb::arg("seeds"), nb::arg("threshold"), nb::arg("negate")=false);
// from asumask.hpp
nb::class_<AsuBrick>(m, "AsuBrick")
.def_ro("size", &AsuBrick::size)
.def_ro("incl", &AsuBrick::incl)
.def("get_extent", &AsuBrick::get_extent)
.def("str", &AsuBrick::str)
;
m.def("find_asu_brick", &find_asu_brick);
}
|
