source: mds-and-trees/tree-genealogy.py @ 564

Last change on this file since 564 was 564, checked in by konrad, 8 years ago

some changes to drawing style + non-random dissimilarity + assertion

File size: 13.2 KB
Line 
1# Draws a genealogical tree (generates a SVG file) based on parent-child relationship information.
2
3import json
4import random
5import math
6import argparse
7
8TIME = "GENERATIONAL"
9BALANCE = "MIN"
10
11# ------SVG---------
12svg_file = 0
13
14svg_line_style = 'stroke="rgb(90%,10%,16%)" stroke-width="1" stroke-opacity="0.8"'
15svg_dot_style = 'r="2" stroke="black" stroke-width="0.2" fill="red"'
16svg_spine_line_style = 'stroke="rgb(0%,0%,80%)" stroke-width="2" stroke-opacity="1"'
17#svg_spine_dot_style = 'r="3" stroke="black" stroke-width="0.4" fill="rgb(50%,50%,100%)"'
18svg_spine_dot_style = 'r="1" stroke="black" stroke-width="0.2" fill="rgb(50%,50%,100%)"'
19
20def svg_add_line(from_pos, to_pos, style=svg_line_style):
21    svg_file.write('<line ' + style + ' x1="' + str(from_pos[0]) + '" x2="' + str(to_pos[0]) + '" y1="' + str(from_pos[1]) + '" y2="' + str(to_pos[1]) + '" />')
22
23def svg_add_dot(pos, style=svg_dot_style):
24    svg_file.write('<circle ' + style + ' cx="' + str(pos[0]) + '" cy="' + str(pos[1]) + '" />')
25
26def svg_generate_line_style(percent):
27    # hotdog
28    from_col = [100, 70, 0]
29    to_col = [60, 0, 0]
30    # neon
31    # from_col = [30, 200, 255]
32    # to_col = [240, 0, 220]
33
34    from_opa = 0.2
35    to_opa = 1.0
36    from_stroke = 1
37    to_stroke = 3
38
39    opa = from_opa*(1-percent) + to_opa*percent
40    stroke = from_stroke*(1-percent) + to_stroke*percent
41
42    percent = 1 - ((1-percent)**20)
43
44    return 'stroke="rgb(' + str(from_col[0]*(1-percent) + to_col[0]*percent) + '%,' \
45           + str(from_col[1]*(1-percent) + to_col[1]*percent) + '%,' \
46           + str(from_col[2]*(1-percent) + to_col[2]*percent) + '%)" stroke-width="' + str(stroke) + '" stroke-opacity="' + str(opa) + '"'
47
48def svg_generate_dot_style(percent):
49    from_col = [100, 90, 0]
50    to_col = [60, 0, 0]
51    # neon
52    # from_col = [30, 200, 255]
53    # to_col = [240, 0, 220]
54
55    from_opa = 0.2
56    to_opa = 1.0
57
58    opa = from_opa*(1-percent) + to_opa*percent
59
60    percent = 1 - ((1-percent)**20)
61
62    return 'fill="rgb(' + str(from_col[0]*(1-percent) + to_col[0]*percent) + '%,' \
63           + str(from_col[1]*(1-percent) + to_col[1]*percent) + '%,' \
64           + str(from_col[2]*(1-percent) + to_col[2]*percent) + '%)" r="1.5" stroke="black" stroke-width="0.2" fill-opacity="' + str(opa) + '" ' \
65           'stroke-opacity="' + str(opa) + '"'
66
67# -------------------
68
69def load_data(dir):
70    global firstnode, nodes, inv_nodes
71    f = open(dir)
72    for line in f:
73        sline = line.split(' ', 2)
74        if len(sline) == 3:
75            if sline[1] == "[OFFSPRING]":
76                creature = json.loads(sline[2])
77                #print("B" +str(creature))
78                if "FromIDs" in creature:
79                    assert(len(creature["FromIDs"]) == 1)
80                    nodes[creature["ID"]] = creature["FromIDs"][0]
81                    if not creature["FromIDs"][0] in nodes:
82                        firstnode = creature["FromIDs"][0]
83
84    for k, v in sorted(nodes.items()):
85        inv_nodes[v] = inv_nodes.get(v, [])
86        inv_nodes[v].append(k)
87
88
89def load_simple_data(dir):
90    global firstnode, nodes, inv_nodes
91    f = open(dir)
92    for line in f:
93        sline = line.split()
94        if len(sline) > 1:
95            #if int(sline[0]) > 15000:
96            #    break
97            if sline[0] == firstnode:
98                continue
99            nodes[sline[0]] = str(max(int(sline[1]), int(firstnode)))
100        else:
101            firstnode = sline[0]
102
103    for k, v in sorted(nodes.items()):
104        inv_nodes[v] = inv_nodes.get(v, [])
105        inv_nodes[v].append(k)
106
107    #print(str(inv_nodes))
108    #quit()
109
110def compute_depth(node):
111    my_depth = 0
112    if node in inv_nodes:
113        for c in inv_nodes[node]:
114            my_depth = max(my_depth, compute_depth(c)+1)
115    depth[node] = my_depth
116    return my_depth
117
118# ------------------------------------
119
120def xmin_crowd(x1, x2, y):
121    if BALANCE == "RANDOM":
122        return (x1 if random.randrange(2) == 0 else x2)
123    elif BALANCE == "MIN":
124        x1_closest = 999999
125        x2_closest = 999999
126        for pos in positions:
127            pos = positions[pos]
128            if pos[1] == y:
129                x1_closest = min(x1_closest, abs(x1-pos[0]))
130                x2_closest = min(x2_closest, abs(x2-pos[0]))
131        return (x1 if x1_closest > x2_closest else x2)
132    elif BALANCE == "DENSITY":
133        x1_dist = 0
134        x2_dist = 0
135        for pos in positions:
136            pos = positions[pos]
137            if pos[1] > y-10 or pos[1] < y+10:
138                dy = pos[1]-y
139                dx1 = pos[0]-x1
140                dx2 = pos[0]-x2
141
142                x1_dist += math.sqrt(dy**2 + dx1**2)
143                x2_dist += math.sqrt(dy**2 + dx2**2)
144        return (x1 if x1_dist > x2_dist else x2)
145
146# ------------------------------------
147
148def prepos_children_reccurent(node):
149    for c in inv_nodes[node]:
150        #print(node + "->" + c)
151        dissimilarity = 0.5 #random.gauss(0,0.3)
152        if TIME == "REAL":
153            id = ""
154            if c[0] == "c":
155                id = int(c[1:])
156            else:
157                id = int(c)
158            positions[c] = [xmin_crowd(positions[node][0]-dissimilarity, positions[node][0]+dissimilarity, id), id]
159        elif TIME == "GENERATIONAL":
160            positions[c] = [xmin_crowd(positions[node][0]-dissimilarity, positions[node][0]+dissimilarity, positions[node][1]+1), positions[node][1]+1]
161
162    for c in inv_nodes[node]:
163        if c in inv_nodes:
164            prepos_children_reccurent(c)
165
166def prepos_children():
167    global max_height, max_width, min_width
168
169    positions[firstnode] = [0, 0]
170
171    prepos_children_reccurent(firstnode)
172
173    for pos in positions:
174        max_height = max(max_height, positions[pos][1])
175        max_width = max(max_width, positions[pos][0])
176        min_width = min(min_width, positions[pos][0])
177
178# ------------------------------------
179
180def draw_children_recurrent(node, max_depth):
181    global max_height, max_width, min_width
182    for c in inv_nodes[node]:
183        if c in inv_nodes:
184            draw_children_recurrent(c, max_depth)
185
186        line_style = (svg_line_style if args.mono_tree else svg_generate_line_style(depth[c]/max_depth))
187        dot_style = (svg_dot_style if args.mono_tree else svg_generate_dot_style(depth[c]/max_depth))
188
189        svg_add_line( (w_margin+w_no_margs*(positions[node][0]-min_width)/(max_width-min_width), h_margin+h_no_margs*positions[node][1]/max_height),
190            (w_margin+w_no_margs*(positions[c][0]-min_width)/(max_width-min_width), h_margin+h_no_margs*positions[c][1]/max_height), line_style)
191        svg_add_dot( (w_margin+w_no_margs*(positions[c][0]-min_width)/(max_width-min_width), h_margin+h_no_margs*positions[c][1]/max_height), dot_style)
192def draw_children():
193    max_depth = 0
194    for k, v in depth.items():
195            max_depth = max(max_depth, v)
196    draw_children_recurrent(firstnode, max_depth)
197    dot_style = (svg_dot_style if args.mono_tree else svg_generate_dot_style(depth[firstnode]/max_depth))
198    svg_add_dot( (w_margin+w_no_margs*(positions[firstnode][0]-min_width)/(max_width-min_width), h_margin+h_no_margs*positions[firstnode][1]/max_height), dot_style)
199
200def draw_spine_recurrent(node):
201    global max_height, max_width, min_width
202    for c in inv_nodes[node]:
203        if depth[c] == depth[node] - 1:
204            if c in inv_nodes:
205                draw_spine_recurrent(c)
206
207            line_style = svg_spine_line_style
208            svg_add_line( (w_margin+w_no_margs*(positions[node][0]-min_width)/(max_width-min_width), h_margin+h_no_margs*positions[node][1]/max_height),
209                (w_margin+w_no_margs*(positions[c][0]-min_width)/(max_width-min_width), h_margin+h_no_margs*positions[c][1]/max_height), line_style)
210            #svg_add_dot( (w_margin+w_no_margs*(positions[c][0]-min_width)/(max_width-min_width), h_margin+h_no_margs*positions[c][1]/max_height), svg_spine_dot_style)
211def draw_spine():
212    draw_spine_recurrent(firstnode)
213    #svg_add_dot( (w_margin+w_no_margs*(positions[firstnode][0]-min_width)/(max_width-min_width), h_margin+h_no_margs*positions[firstnode][1]/max_height), svg_spine_dot_style)
214
215def draw_skeleton_reccurent(node, max_depth):
216    global max_height, max_width, min_width
217    for c in inv_nodes[node]:
218        if depth[c] >= min_skeleton_depth or depth[c] == max([depth[q] for q in inv_nodes[node]]):
219            if c in inv_nodes:
220                draw_skeleton_reccurent(c, max_depth)
221
222            line_style = svg_spine_line_style
223            svg_add_line( (w_margin+w_no_margs*(positions[node][0]-min_width)/(max_width-min_width), h_margin+h_no_margs*positions[node][1]/max_height),
224                (w_margin+w_no_margs*(positions[c][0]-min_width)/(max_width-min_width), h_margin+h_no_margs*positions[c][1]/max_height), line_style)
225            #svg_add_dot( (w_margin+w_no_margs*(positions[c][0]-min_width)/(max_width-min_width), h_margin+h_no_margs*positions[c][1]/max_height),
226            #             svg_spine_dot_style)
227def draw_skeleton():
228    max_depth = 0
229    for k, v in depth.items():
230            max_depth = max(max_depth, v)
231
232    draw_skeleton_reccurent(firstnode, max_depth)
233    #svg_add_dot( (w_margin+w_no_margs*(positions[firstnode][0]-min_width)/(max_width-min_width), h_margin+h_no_margs*positions[firstnode][1]/max_height),
234    #             svg_spine_dot_style)
235
236
237
238##################################################### main #####################################################
239
240args = 0
241
242h = 800
243w = 600
244h_margin = 10
245w_margin = 10
246h_no_margs = h - 2* h_margin
247w_no_margs = w - 2* w_margin
248
249max_height = 0
250max_width = 0
251min_width = 9999999999
252
253min_skeleton_depth = 0
254
255firstnode = ""
256nodes = {}
257inv_nodes = {}
258positions = {}
259depth = {}
260
261def main():
262    global svg_file, min_skeleton_depth, args, TIME, BALANCE
263
264    parser = argparse.ArgumentParser(description='Process some integers.')
265    parser.add_argument('--in', dest='input', required=True, help='input file with stuctured evolutionary data')
266    parser.add_argument('--out', dest='output', required=True, help='output file for the evolutionary tree')
267    draw_tree_parser = parser.add_mutually_exclusive_group(required=False)
268    draw_tree_parser.add_argument('--draw-tree', dest='draw_tree', action='store_true', help='whether drawing the full tree should be skipped')
269    draw_tree_parser.add_argument('--no-draw-tree', dest='draw_tree', action='store_false')
270
271    draw_skeleton_parser = parser.add_mutually_exclusive_group(required=False)
272    draw_skeleton_parser.add_argument('--draw-skeleton', dest='draw_skeleton', action='store_true', help='whether the skeleton of the tree should be drawn')
273    draw_skeleton_parser.add_argument('--no-draw-skeleton', dest='draw_skeleton', action='store_false')
274
275    draw_spine_parser = parser.add_mutually_exclusive_group(required=False)
276    draw_spine_parser.add_argument('--draw-spine', dest='draw_spine', action='store_true', help='whether the spine of the tree should be drawn')
277    draw_spine_parser.add_argument('--no-draw-spine', dest='draw_spine', action='store_false')
278
279    #TODO: better names for those parameters
280    parser.add_argument('--time', default='REAL', dest='time', help='values on vertical axis (REAL/GENERATIONAL)')
281    parser.add_argument('--balance', default='MIN',dest='balance', help='method of placing node in the tree (RANDOM/MIN/DENSITY)')
282
283    mono_tree_parser = parser.add_mutually_exclusive_group(required=False)
284    mono_tree_parser.add_argument('--mono-tree', dest='mono_tree', action='store_true', help='whether the tree should be drawn with a single color')
285    mono_tree_parser.add_argument('--no-mono-tree', dest='mono_tree', action='store_false')
286
287    parser.add_argument('--min-skeleton-depth', type=int, default=2, dest='min_skeleton_depth', help='minimal distance from the leafs for the nodes in the skeleton')
288    parser.add_argument('--seed', type=int, dest='seed', help='seed for the random number generator (-1 for random)')
289
290    parser.add_argument('--simple-data', type=bool, dest='simple_data', help='input data are given in a simple format (#child #parent)')
291
292    parser.set_defaults(mono_tree=False)
293    parser.set_defaults(draw_tree=True)
294    parser.set_defaults(draw_skeleton=False)
295    parser.set_defaults(draw_spine=False)
296
297    parser.set_defaults(seed=-1)
298
299    args = parser.parse_args()
300
301    TIME = args.time
302    BALANCE = args.balance
303
304    dir = args.input
305    min_skeleton_depth = args.min_skeleton_depth
306    seed = args.seed
307    if seed == -1:
308        seed = random.randint(0, 10000)
309    random.seed(seed)
310    print("seed:", seed)
311
312    if args.simple_data:
313        load_simple_data(dir)
314    else:
315        load_data(dir)
316
317    compute_depth(firstnode)
318
319    svg_file = open(args.output, "w")
320    svg_file.write('<svg xmlns:svg="http://www.w3.org/2000/svg" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" version="1.0" '
321                   'width="' + str(w) + '" height="' + str(h) + '">')
322
323    prepos_children()
324
325    if args.draw_tree:
326        draw_children()
327    if args.draw_skeleton:
328        draw_skeleton()
329    if args.draw_spine:
330        draw_spine()
331
332    svg_file.write("</svg>")
333    svg_file.close()
334
335main()
336
Note: See TracBrowser for help on using the repository browser.