summaryrefslogtreecommitdiffstats
path: root/src/interpolation.py
blob: 6c2bce1550d3490acc2c9c8d5bceb040a759b154 (plain) (blame)
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
#!/usr/bin/env python3

# Copyright © 2014  Mattias Andrée (maandree@member.fsf.org)
# 
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
# 
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
# 
# You should have received a copy of the GNU General Public License
# along with this program.  If not, see <http://www.gnu.org/licenses/>.

# This module contains interpolation functions.

from aux import *
from curve import *


def linearly_interpolate_ramp(r, g, b):
    '''
    Linearly interpolate ramps to the size of the output axes
    
    @param   r:list<float>                                   The red colour curves
    @param   g:list<float>                                   The green colour curves
    @param   b:list<float>                                   The blue colour curves
    @return  :(r:list<float>, g:list<float>, b:list<float>)  The input parameters extended to sizes of `o_size`,
                                                             or their original size, whatever is larger.
    '''
    C = lambda c : c[:] if len(c) >= o_size else ([None] * o_size)
    R, G, B = C(r), C(g), C(b)
    for small, large in ((r, R), (g, G), (b, B)):
        small_, large_ = len(small) - 1, len(large) - 1
        # Only interpolate if scaling up
        if large_ > small_:
            for i in range(len(large)):
                # Scaling
                j = i * small_ / large_
                # Floor, weight, ceiling
                j, w, k = int(j), j % 1, min(int(j) + 1, small_)
                # Interpolation
                large[i] = small[j] * (1 - w) + small[k] * w
    return (R, G, B)


def cubicly_interpolate_ramp(r, g, b, tension = 0):
    '''
    Interpolate ramps to the size of the output axes using cubic Hermite spline
    
    @param   r:list<float>                                   The red colour curves
    @param   g:list<float>                                   The green colour curves
    @param   b:list<float>                                   The blue colour curves
    @param   tension:float                                   A [0, 1] value of the tension
    @return  :(r:list<float>, g:list<float>, b:list<float>)  The input parameters extended to sizes of `o_size`,
                                                             or their original size, whatever is larger.
    '''
    C = lambda c : c[:] if len(c) >= o_size else ([None] * o_size)
    R, G, B = C(r), C(g), C(b)
    # Basis functions
    #h00 = lambda t : (1 + 2 * t) * (1 - t) ** 2
    h10 = lambda t : t * (1 - t) ** 2
    h01 = lambda t : t ** 2 * (3 - 2 * t)
    h11 = lambda t : t ** 2 * (t - 1)
    def tangent(values, index, last):
        '''
        Calculate the tangent at a point
        
        @param   values:list<float>  Mapping from points to values
        @param   index:int           The point
        @param   last:int            The last point
        @return  :float              The tangent at the point `index`
        '''
        if last == 0:      return 0
        if index == 0:     return values[1] - values[0]
        if index == last:  return values[last] - values[last - 1]
        return (values[index + 1] - values[index - 1]) / 2
    # Tension coefficent
    c_ = 1 - tension
    # Interpolate each curve
    for small, large in ((r, R), (g, G), (b, B)):
        small_, large_ = len(small) - 1, len(large) - 1
        # Only interpolate if scaling up
        if large_ > small_:
            for i in range(len(large)):
                # Scaling
                j = i * small_ / large_
                # Floor, weight, ceiling
                j, w, k = int(j), j % 1, min(int(j) + 1, small_)
                # Points
                pj, pk = small[j], small[k]
                # Tangents
                mj, mk = c_ * tangent(small, j, small_), c_ * tangent(small, k, small_)
                # Interpolation
                large[i] = pj + h10(w) * mj + h01(w) * (pk - pj) + h11(w) * mk
    ## Check local monotonicity
    eliminate_halos(r, g, b, R, G, B)
    return (R, G, B)


def monotonic_cubicly_interpolate_ramp(r, g, b, tension = 0):
    '''
    Interpolate ramps to the size of the output axes using
    monotone cubic Hermite spline and the Fritsch–Carlson method
    
    Does not overshoot, but regular cubic interpolation with uses
    linear replacement for overshot areas is better
    
    @param   r:list<float>                                   The red colour curves
    @param   g:list<float>                                   The green colour curves
    @param   b:list<float>                                   The blue colour curves
    @param   tension:float                                   A [0, 1] value of the tension
    @return  :(r:list<float>, g:list<float>, b:list<float>)  The input parameters extended to sizes of `o_size`,
                                                             or their original size, whatever is larger.
    '''
    C = lambda c : c[:] if len(c) >= o_size else ([None] * o_size)
    R, G, B = C(r), C(g), C(b)
    # Basis functions
    #h00 = lambda t : (1 + 2 * t) * (1 - t) ** 2
    h10 = lambda t : t * (1 - t) ** 2
    h01 = lambda t : t ** 2 * (3 - 2 * t)
    h11 = lambda t : t ** 2 * (t - 1)
    def tangent(values, index, last):
        '''
        Calculate the tangent at a point
        
        @param   values:list<float>  Mapping from points to values
        @param   index:int           The point
        @param   last:int            The last point
        @return  :float              The tangent at the point `index`
        '''
        if last == 0:      return 0
        if index == 0:     return values[1] - values[0]
        if index == last:  return values[last] - values[last - 1]
        return (values[index + 1] - values[index - 1]) / 2
    # Tension coefficent
    c_ = 1 - tension
    ## Interpolant selection
    # Compute the slopes of the secant
    # lines between successive points
    ds = [small[i + 1] - small[i] for i in range(small_)]
    # Initialize the tangents at every
    # data point as the average of the secants
    ms = [ds[0]] + [(ds[i - 1] + ds[i]) / 2 for i in range(1, small_)] + [ds[small_ - 1]]
    βlast = 0
    for i in range(small_):
        if ds[i] == 0:
            # Two successive values are equal, ms[i],
            # must be zero to preserve monotonicity,
            # no idea to do further work on them.
            ms[i], βlast = 0, -1
            continue
        # Look for local extremums
        α, β = ms[i] / ds[i], ms[i + 1] / ds[i]
        if (α < 0) or (βlast < 0):
            # Local extremum found,
            # ensure piecewise monotonicity
            ms[i], β = 0, -1
        elif α ** 2 + β ** 2 > 9:
            # Otherwise, prevent overshoot and ensure
            # monotonicity by restricting the (α, β)
            # vector to a circle of radius 3.
            τ = 3 / (α ** 2 + β ** 2) ** 0.5
            ms[i], ms[i + 1] = τ * α * ds[i], τ * β  * ds[i]
        βlast = β
    ## Interpolate each curve
    for small, large in ((r, R), (g, G), (b, B)):
        small_, large_ = len(small) - 1, len(large) - 1
        # Only interpolate if scaling up
        if large_ > small_:
            for i in range(len(large)):
                # Scaling
                j = i * small_ / large_
                # Floor, weight, ceiling
                j, w, k = int(j), j % 1, min(int(j) + 1, small_)
                # Points
                pj, pk = small[j], small[k]
                # Tangents
                mj, mk = c_ * ms[j], c_ * ms[k]
                # Interpolation
                large[i] = pj + h10(w) * mj + h01(w) * (pk - pj) + h11(w) * mk
    ## Check local monotonicity
    eliminate_halos(r, g, b, R, G, B)
    return (R, G, B)


def polynomially_interpolate_ramp(r, g, b): # TODO Speedup, demo and document this
    '''
    Polynomially interpolate ramps to the size of the output axes.
    
    This function will replace parts of the result with linear interpolation
    where local monotonicity have been broken. That is, there is a local
    maximum or local minimum generated between two reference points, linear
    interpolation will be used instead between those two points.
    
    @param   r:list<float>                                   The red colour curves
    @param   g:list<float>                                   The green colour curves
    @param   b:list<float>                                   The blue colour curves
    @return  :(r:list<float>, g:list<float>, b:list<float>)  The input parameters extended to sizes of `o_size`,
                                                             or their original size, whatever is larger.
    '''
    C = lambda c : c[:] if len(c) >= o_size else ([None] * o_size)
    R, G, B, linear = C(r), C(g), C(b), [None]
    for ci, (small, large) in enumerate(((r, R), (g, G), (b, B))):
        small_, large_ = len(small) - 1, len(large) - 1
        # Only interpolate if scaling up
        if large_ > small_:
            n = len(small)
            ## Construct interpolation matrix (TODO this is not working correctly)
            M = [[small[y] ** i for i in range(n)] for y in range(n)]
            A = [x / small_ for x in range(n)]
            ## Eliminate interpolation matrix
            # (XXX this can be done faster by utilising the fact that we have a Vandermonde matrix)
            # Eliminiate lower left
            for k in range(n - 1):
                for i in range(k + 1, n):
                    m = M[i][k] / M[k][k]
                    M[i][k + 1:] = [M[i][j] - M[k][j] * m for j in range(k + 1, n)]
                    A[i] -= A[k] * m
            # Eliminiate upper right
            for k in reversed(range(n)):
                A[:k] = [A[i] - A[k] * M[i][k] / M[k][k] for i in range(k)]
            # Eliminiate diagonal
            A = [A[k] / M[k][k] for k in range(n)]
            ## Construct interpolation function
            f = lambda x : sum(A[i] * x ** i for i in range(n))
            ## Apply interpolation
            large[:] = [f(x / large_) for x in range(len(large))]
    ## Check local monotonicity
    eliminate_halos(r, g, b, R, G, B)
    return (R, G, B)


def eliminate_halos(r, g, b, R, G, B):
    '''
    Eliminate haloing effects in interpolations
    
    @param  r:list<float>  The original red curve
    @param  g:list<float>  The original green curve
    @param  b:list<float>  The original blue curve
    @param  R:list<float>  The scaled up red curve
    @param  G:list<float>  The scaled up green curve
    @param  B:list<float>  The scaled up blue curve
    '''
    linear = None
    for ci, (small, large) in enumerate(((r, R), (g, G), (b, B))):
        small_, large_ = len(small) - 1, len(large) - 1
        ## Check local monotonicity
        for i in range(small_):
            # Small curve
            x1, x2, y1, y2 = i, i + 1, small[i], small[i + 1]
            # Scaled up curve
            X1, X2 = int(x1 * large_ / small_), int(x2 * large_ / small_)
            Y1, Y2 = large[X1], large[X2]
            monotone = True
            if y2 == y1:
                # Flat part, just make sure it is flat in the interpolation
                # without doing a check before.
                for x in range(X1, X2 + 1):
                    large[x] = y1
            elif y2 > y1:
                # Increasing
                monotone = all(map(lambda x : large[x + 1] >= large[x], range(X1, X2))) and (Y2 > Y1)
            elif y2 < y1:
                # Decreasing
                monotone = all(map(lambda x : large[x + 1] <= large[x], range(X1, X2))) and (Y2 < Y1)
            # If the monotonicity has been broken,
            if not monotone:
                # replace the partition with linear interpolation.
                # If linear interpolation has not yet been calculated,
                if linear is None:
                    # then calculate it.
                    linear = linearly_interpolate_ramp(r, g, b)
                # Extract the linear interpolation for the current colour curve,
                # and replace the local partition with the linear interpolation.
                large[X1 : X2 + 1] = linear[ci][X1 : X2 + 1]


def interpolate_function(function, interpolator):
    '''
    Interpolate a function that applies adjustments from a lookup table
    
    @param   function:()→void                                 The function that applies the adjustments
    @param   interpolator:(list<float>{3})?→[list<float>{3}]  Function that interpolates lookup tables
    @return  :()→void                                         `function` interpolated
    '''
    # Do not interpolation if none is selected
    if interpolator is None:
        return function
    # Store the current adjustments, we
    # will need to apply our own temporary
    # adjustments
    stored = store()
    # Clean any adjustments,
    start_over()
    # and apply those we should interpolate.
    function()
    # Interpolate the adjustments we just
    # made and make a function out of it
    rc = functionise(interpolator(*store()))
    # Restore the adjustments to those
    # that were applied when we started
    restore(stored)
    return rc