-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathplot_results.jl
253 lines (209 loc) · 9 KB
/
plot_results.jl
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
@time using Glob
@time using MAT
using JSON
@time using Plots
# @time using StatsPlots
using CurveFit
import Unmarshal
include("base_structs.jl")
function angle_to_color(degree, sv=1.0)
color = convert(RGB, HSV(degree, 1.0, sv))
return color
end
function fix_globs(file_list)
new_file_list = [ "" ]
pop!(new_file_list)
for i = 1 : length(file_list)
file_str = file_list[ i ]
push!(new_file_list, replace(file_str, "\\" => "/" ))
end
return new_file_list
end
function compute_avg(arr)
return (1.0 * sum(arr)) / length(arr)
end
function compute_var(arr)
avg = compute_avg(arr)
var = 0.0
for i = 1 : length(arr)
var += (arr[ i ] - avg)^2
end
return var / length(arr)
end
function compute_std(arr)
var = compute_var(arr)
return sqrt(var)
end
function reverse_dict(some_dict)
all_keys = keys(some_dict)
(nkeys, first_key) = iterate(all_keys)
println(typeof(nkeys))
println(typeof(first_key))
println(nkeys)
println(first_key)
println(some_dict[nkeys])
new_dict = Dict([ (some_dict[ nkeys ], nkeys) ])
for key in all_keys
new_dict[ some_dict[ key ] ] = key
end
return new_dict
end
function comp_avg(list_of_vals)
sum_vals = 0.0
if length(list_of_vals) == 0
return -1
end
for i = 1 : length(list_of_vals)
sum_vals += list_of_vals[ i ]
end
return sum_vals / length(list_of_vals)
end
function comp_med(list_of_vals)
med_ind = Int(ceil(length(list_of_vals) / 2))
sl = sort(list_of_vals)
return sl[ med_ind ]
end
function comp_lowquant(list_of_vals)
low_ind = Int(floor(length(list_of_vals) / 4))
sl = sort(list_of_vals)
if low_ind == 0
low_ind = 1
end
return sl[ low_ind ]
end
function comp_upquant(list_of_vals)
up_ind = Int(ceil(3 * length(list_of_vals) / 4))
sl = sort(list_of_vals)
return sl[ up_ind ]
end
function gen_plots(train_nbits, clen, mclauses, kinsts, pdepth, DPI=100)
train_ut_dir = string("./train_angles", "_ut", "_clen=", clen, "/")
train_str = string("rand_insts", "_nbits=", train_nbits, "_mclauses=", mclauses, "_kinsts=", kinsts, "_pdepth=", pdepth)
train_path = string(train_ut_dir, train_str, ".json")
nbit_range = [ 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 ]
### SET UP LOAD/SAVE ###
ut_dir = string("./results_qaoa_ut_clen=", clen ,"_pdepth=", pdepth, "/")
if !isdir(ut_dir)
mkdir(ut_dir)
end
succ_vs_nbit_per_inst = [ Array{Float64, 1}() for i = 1 : length(nbit_range) ]
low_succ_vs_nbit_per_inst = [ 0.0 for i = 1 : length(nbit_range) ]
med_succ_vs_nbit_per_inst = [ 0.0 for i = 1 : length(nbit_range) ]
upp_succ_vs_nbit_per_inst = [ 0.0 for i = 1 : length(nbit_range) ]
ut_low_size_vs_nbit_per_inst = [ 0.0 for i = 1 : length(nbit_range) ]
ut_med_size_vs_nbit_per_inst = [ 0.0 for i = 1 : length(nbit_range) ]
ut_upp_size_vs_nbit_per_inst = [ 0.0 for i = 1 : length(nbit_range) ]
flatten_nbits_per_inst = Array{Float64, 1}()
flatten_succ_per_inst = Array{Float64, 1}()
for nbit_id = 1 : length(nbit_range)
nbits = nbit_range[nbit_id]
search_str1 = string(ut_dir, "inst_nbits=", nbits, "/*.json")
all_file_strs = fix_globs(glob(search_str1))
# println(all_file_strs)
size_per_inst = Array{Float64, 1}()
succ_per_inst = Array{Float64, 1}()
for i = 1 : length(all_file_strs)
open(all_file_strs[ i ], "r") do f
json_string = JSON.read(f, String)
new_res = Unmarshal.unmarshal(ExpResults, JSON.parse(json_string))
push!(succ_per_inst, new_res.end_sup)
push!(size_per_inst, new_res.num_states)
push!(flatten_succ_per_inst, new_res.end_sup)
push!(flatten_nbits_per_inst, nbits)
end
end
low_succ_vs_nbit_per_inst[ nbit_id ] = comp_lowquant(succ_per_inst)
med_succ_vs_nbit_per_inst[ nbit_id ] = comp_med(succ_per_inst)
upp_succ_vs_nbit_per_inst[ nbit_id ] = comp_upquant(succ_per_inst)
ut_low_size_vs_nbit_per_inst[ nbit_id ] = comp_lowquant(size_per_inst)
ut_med_size_vs_nbit_per_inst[ nbit_id ] = comp_med(size_per_inst)
ut_upp_size_vs_nbit_per_inst[ nbit_id ] = comp_upquant(size_per_inst)
end
plot_succ_vs_nbits = Plots.plot(legend=:bottomright, title="QAOA vs. tQAOA on 1-in-3 SAT",
xlabel="Number of Variables", ylabel="Log of Success Prob")
Plots.scatter!( flatten_nbits_per_inst, log.(flatten_succ_per_inst), label=string("QAOA"),
seriescolor=angle_to_color(360*(1/3)), markersize=3, markeralpha=0.1, markershape=:cross)
lsig = [ abs(med_succ_vs_nbit_per_inst[i] - low_succ_vs_nbit_per_inst[i]) + 0.001 for i = 1 : length(med_succ_vs_nbit_per_inst) ]
usig = [ abs(upp_succ_vs_nbit_per_inst[i] - med_succ_vs_nbit_per_inst[i]) + 0.001 for i = 1 : length(med_succ_vs_nbit_per_inst) ]
Plots.plot!(nbit_range, log.(med_succ_vs_nbit_per_inst), label="", linewidth=4, linestyle=:solid,
seriescolor=angle_to_color(360*(1/3)), ribbon=(lsig, usig), fillalpha=0.1)
exp_fit = curve_fit(ExpFit, flatten_nbits_per_inst, flatten_succ_per_inst)
println(exp_fit)
train_t_dir = string("./train_angles", "_t", "_clen=", clen, "/")
train_str = string("rand_insts", "_nbits=", train_nbits, "_mclauses=", mclauses, "_kinsts=", kinsts, "_pdepth=", pdepth)
train_path = string(train_ut_dir, train_str, ".json")
nbit_range = [ 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 ]
### SET UP LOAD/SAVE ###
t_dir = string("./results_qaoa_t_clen=", clen ,"_pdepth=", pdepth, "/")
if !isdir(t_dir)
mkdir(t_dir)
end
succ_vs_nbit_per_inst = [ Array{Float64, 1}() for i = 1 : length(nbit_range) ]
t_low_size_vs_nbit_per_inst = [ 0.0 for i = 1 : length(nbit_range) ]
t_med_size_vs_nbit_per_inst = [ 0.0 for i = 1 : length(nbit_range) ]
t_upp_size_vs_nbit_per_inst = [ 0.0 for i = 1 : length(nbit_range) ]
flatten_nbits_per_inst = Array{Float64, 1}()
flatten_succ_per_inst = Array{Float64, 1}()
for nbit_id = 1 : length(nbit_range)
nbits = nbit_range[nbit_id]
search_str1 = string(t_dir, "inst_nbits=", nbits, "/*.json")
all_file_strs = fix_globs(glob(search_str1))
# println(all_file_strs)
size_per_inst = Array{Float64, 1}()
succ_per_inst = Array{Float64, 1}()
for i = 1 : length(all_file_strs)
open(all_file_strs[ i ], "r") do f
json_string = JSON.read(f, String)
new_res = Unmarshal.unmarshal(ExpResults, JSON.parse(json_string))
push!(succ_per_inst, new_res.end_sup)
push!(size_per_inst, new_res.num_states)
push!(flatten_succ_per_inst, new_res.end_sup)
push!(flatten_nbits_per_inst, nbits)
end
end
low_succ_vs_nbit_per_inst[ nbit_id ] = comp_lowquant(succ_per_inst)
med_succ_vs_nbit_per_inst[ nbit_id ] = comp_med(succ_per_inst)
upp_succ_vs_nbit_per_inst[ nbit_id ] = comp_upquant(succ_per_inst)
t_low_size_vs_nbit_per_inst[ nbit_id ] = comp_lowquant(size_per_inst)
t_med_size_vs_nbit_per_inst[ nbit_id ] = comp_med(size_per_inst)
t_upp_size_vs_nbit_per_inst[ nbit_id ] = comp_upquant(size_per_inst)
end
Plots.scatter!( flatten_nbits_per_inst, log.(flatten_succ_per_inst), label=string("tQAOA"),
seriescolor=angle_to_color(360*(2/3)), markersize=3, markeralpha=0.1, markershape=:cross)
lsig = [ abs(med_succ_vs_nbit_per_inst[i] - low_succ_vs_nbit_per_inst[i]) + 0.001 for i = 1 : length(med_succ_vs_nbit_per_inst) ]
usig = [ abs(upp_succ_vs_nbit_per_inst[i] - med_succ_vs_nbit_per_inst[i]) + 0.001 for i = 1 : length(med_succ_vs_nbit_per_inst) ]
Plots.plot!(nbit_range, log.(med_succ_vs_nbit_per_inst), label="", linewidth=4, linestyle=:solid,
seriescolor=angle_to_color(360*(2/3)), ribbon=(lsig, usig), fillalpha=0.1)
exp_fit2 = curve_fit(ExpFit, flatten_nbits_per_inst, flatten_succ_per_inst)
println(exp_fit2)
Plots.plot!(ylim=(-0.5,0.0), legend=:bottomleft)
if pdepth == 60
Plots.plot!(ylim=(-0.1,0.0), legend=:bottomleft)
end
Plots.plot!(dpi=DPI)
DO_SAVE = 1
if DO_SAVE == 1
savefig(plot_succ_vs_nbits, string("./plots/", "plot_succ_vs_nbits_pdepth=", pdepth, ".png"))
end
#=
plot_size_vs_nbits = Plots.plot(legend=:bottomright, title="QAOA vs. tQAOA on 1-in-3 SAT",
xlabel="Number of Variables", ylabel="Log of Search Space Dim")
utlsig = [ abs(ut_med_size_vs_nbit_per_inst[i] - ut_low_size_vs_nbit_per_inst[i]) + 0.001 for i = 1 : length(ut_med_size_vs_nbit_per_inst) ]
utusig = [ abs(ut_upp_size_vs_nbit_per_inst[i] - ut_med_size_vs_nbit_per_inst[i]) + 0.001 for i = 1 : length(ut_med_size_vs_nbit_per_inst) ]
Plots.plot!(nbit_range, ut_med_size_vs_nbit_per_inst, label="QAOA", linewidth=4, linestyle=:solid,
seriescolor=angle_to_color(360*(1/3)), ribbon=(utlsig, utusig), fillalpha=0.1)
tlsig = [ abs(t_med_size_vs_nbit_per_inst[i] - t_low_size_vs_nbit_per_inst[i]) + 0.001 for i = 1 : length(t_med_size_vs_nbit_per_inst) ]
tusig = [ abs(t_upp_size_vs_nbit_per_inst[i] - t_med_size_vs_nbit_per_inst[i]) + 0.001 for i = 1 : length(t_med_size_vs_nbit_per_inst) ]
Plots.plot!(nbit_range, log.(t_med_size_vs_nbit_per_inst), label="tQAOA", linewidth=4, linestyle=:solid,
seriescolor=angle_to_color(360*(2/3)), ribbon=(tlsig, tusig), fillalpha=0.1)
Plots.plot!(dpi=DPI)
DO_SAVE = 1
if DO_SAVE == 1
savefig(plot_size_vs_nbits, string("./plots/", "plot_size_vs_nbits.png"))
end
=#
end
gen_plots(12, 3, 4, 100, 14, 300)
gen_plots(12, 3, 4, 100, 60, 300)
# gen_plots(12, 3, 4, 100, 14, 100)
# gen_plots(12, 3, 4, 100, 60, 100)