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
|
#!/usr/bin/python3
# -*- coding: utf-8 -*-
# Copyright (C) 2009-2020 Authors of CryptoMiniSat, see AUTHORS file
#
# 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; version 2
# of the License.
#
# 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, write to the Free Software
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
# 02110-1301, USA.
import numpy as np
import pandas as pd
import xgboost as xgb
import os
from ccg import *
# to test memory safety
#import gc
# to test memory usage
#from memory_profiler import *
MISSING=np.NaN
raw_data = [
"is_ternary_resolvent",
"rdb0.which_red_array",
"rdb0.last_touched",
"rdb0.act_ranking_rel",
"rdb0.uip1_ranking_rel",
"rdb0.prop_ranking_rel",
"rdb0.last_touched_diff",
"cl.time_inside_solver",
"rdb0.props_made",
"rdb0_common.avg_props",
#"rdb0_common.avg_glue", CANNOT DO, ternaries have no glue!
"rdb0_common.avg_uip1_used",
"rdb0_common.conflSizeHistLT_avg",
"rdb0_common.glueHistLT_avg",
"rdb0.sum_props_made",
"rdb0.discounted_props_made",
"rdb0.discounted_props_made2",
"rdb0.discounted_props_made3",
"rdb0.discounted_uip1_used",
"rdb0.discounted_uip1_used2",
"rdb0.discounted_uip1_used3",
"rdb0.sum_uip1_used",
"rdb0.uip1_used",
"rdb0.size",
"rdb0.sum_uip1_per_time_ranking",
"rdb0.sum_props_per_time_ranking",
"rdb0.sum_uip1_per_time_ranking_rel",
"rdb0.sum_props_per_time_ranking_rel",
"rdb0.is_distilled",
"cl.glueHist_avg",
"cl.atedecents_binIrred",
"cl.glueHistLT_avg",
"cl.glueHist_longterm_avg",
"cl.num_antecedents",
"cl.overlapHistLT_avg",
"cl.conflSizeHist_avg",
"cl.atedecents_binRed",
"cl.num_total_lits_antecedents",
"cl.numResolutionsHistLT_avg",
"rdb0.glue",
"cl.orig_glue",
"cl.glue_before_minim",
"cl.trail_depth_level",
#"sum_uip1_per_time_ranking_rel",
#"sum_props_per_time_ranking_rel",
]
# check reproducibility by dumping and checking against previous run's dump
def dump_or_check(fname, df):
if check_file_exists(fname):
df_saved = pd.read_pickle(fname)
print("Checking equals...", fname)
if not df.equals(df_saved):
print("df.shape :", df.shape)
print("df_saved.shape :", df_saved.shape)
assert df.shape == df_saved.shape
for i in range(df):
if df[i] != df_saved[i]:
print("Not equal:")
print(df[i])
print(df_saved[i])
exit(-1)
else:
df.to_pickle(fname)
print("Not checking, writing: ", fname)
def check_file_exists(fname):
return os.path.exists(fname)
def get_features(fname):
best_features = []
if not check_file_exists(fname):
print("File '%s' not accessible" % fname)
exit(-1)
with open(fname, "r") as f:
for l in f:
l = l.strip()
if len(l) == 0:
continue
if l[0] == "#":
continue
best_features.append(l)
return best_features
def check_against_binary_dat(fname, df, df_raw):
global best_features
check_file_exists(fname)
df2 = pd.read_csv(fname, sep=",", names=best_features)
print("Checking binary vs python:", fname)
for f in best_features:
for i in range(df.shape[0]):
s_pyt = float(df[f][i])
s_bin = float(df2[f][i])
if np.isnan(s_pyt) and np.isnan(s_bin):
continue
if np.isnan(s_pyt) and not np.isnan(s_bin):
assert False
if not np.isnan(s_pyt) and np.isnan(s_bin):
assert False
diff = abs(s_pyt - s_bin)
if diff > 10e-5:
for f_raw in list(df_raw):
print("pyt_raw:", df_raw[f_raw][i], " feat: ", f_raw)
print("pyt:", s_pyt, " feat: ", f)
print("bin:", s_bin, " feat: ", f)
print("diff for feat ", f, ": ", diff)
else:
#print("OK for feat", f)
pass
assert diff < 10e-5
#assert df.equals(df2)
models = []
best_features = []
feat_gen_exprs = []
feat_gen_funcs = []
def add_features(df, df2):
for i, feat_gen_func in zip(range(len(best_features)), feat_gen_funcs):
df2[:, i] = feat_gen_func(df)
def set_up_features(features_fname):
global best_features
global feat_gen_exprs
global feat_gen_funcs
best_features = get_features(features_fname)
for i, feat in zip(range(len(best_features)), best_features):
feat_gen_expr = ccg.to_source(ast.parse(feat))
feat_gen_exprs.append(feat_gen_expr)
create_function = "def a%d(df): return %s" % (i, feat_gen_expr)
exec(create_function)
exec("feat_gen_funcs.append(a%d)" % i)
#print(feat_gen_funcs)
def load_models(short_fname, long_fname, forever_fname):
global models
for fname in [short_fname, long_fname, forever_fname]:
clf_xgboost = xgb.XGBRegressor(n_jobs=1)
new_fname = fname.replace("-py.", "-xgb.")
if not os.path.exists(new_fname):
new_fname = fname.replace("-py.", ".")
clf_xgboost.load_model(new_fname)
models.append(clf_xgboost)
num_called = 0
# to test memory usage
#@profile
def predict(data, check=False, dump=False):
global num_called
ret = []
df = pd.DataFrame(data, columns=raw_data)
transformed_data = np.empty((df.shape[0], len(best_features)), dtype=float)
if check:
dump_or_check('df_check'+str(num_called), df)
add_features(df, transformed_data)
df_final = pd.DataFrame(transformed_data, columns=best_features)
df_final.replace([np.inf, np.NaN, np.inf, np.NINF, np.Infinity], MISSING, inplace=True)
if check:
check_against_binary_dat('bin_dump'+str(num_called)+".csv", df_final, df)
for i in range(3):
#x = models[i].predict(df_final)
x = models[i].get_booster().inplace_predict(df_final)
if check:
dump_or_check('x-%d-%d' % (num_called, i), pd.DataFrame(x))
ret.append(x)
if dump:
for i,name in zip(range(3), ["short", "long", "forever"]):
df["x.used_later_%s" % name] = ret[i]
df.to_pickle('df_dump_%s' % num_called)
if check:
dump_or_check('df_pred'+str(num_called), df_final)
num_called += 1
del df
del df_final
del transformed_data
# gc.collect() # to test memory safety
return ret
|