line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
1
|
|
|
|
|
|
|
|
2
|
|
|
|
|
|
|
#==================================================================== -*-perl-*- |
3
|
|
|
|
|
|
|
# |
4
|
|
|
|
|
|
|
# DateTime::Event::Predict |
5
|
|
|
|
|
|
|
# |
6
|
|
|
|
|
|
|
# DESCRIPTION |
7
|
|
|
|
|
|
|
# Predict new dates from a set of dates |
8
|
|
|
|
|
|
|
# |
9
|
|
|
|
|
|
|
# AUTHORS |
10
|
|
|
|
|
|
|
# Brian Hann |
11
|
|
|
|
|
|
|
# |
12
|
|
|
|
|
|
|
#=============================================================================== |
13
|
|
|
|
|
|
|
|
14
|
|
|
|
|
|
|
package DateTime::Event::Predict; |
15
|
|
|
|
|
|
|
|
16
|
2
|
|
|
2
|
|
71435
|
use 5.006; |
|
2
|
|
|
|
|
9
|
|
|
2
|
|
|
|
|
73
|
|
17
|
|
|
|
|
|
|
|
18
|
2
|
|
|
2
|
|
11
|
use strict; |
|
2
|
|
|
|
|
3
|
|
|
2
|
|
|
|
|
76
|
|
19
|
|
|
|
|
|
|
|
20
|
2
|
|
|
2
|
|
2249
|
use DateTime; |
|
0
|
|
|
|
|
|
|
|
0
|
|
|
|
|
|
|
21
|
|
|
|
|
|
|
use Params::Validate qw(:all); |
22
|
|
|
|
|
|
|
use Carp qw(carp croak confess); |
23
|
|
|
|
|
|
|
use Scalar::Util; |
24
|
|
|
|
|
|
|
use Data::Dumper; |
25
|
|
|
|
|
|
|
|
26
|
|
|
|
|
|
|
use POSIX qw(ceil); |
27
|
|
|
|
|
|
|
|
28
|
|
|
|
|
|
|
use DateTime::Event::Predict::Profile qw(:buckets); |
29
|
|
|
|
|
|
|
|
30
|
|
|
|
|
|
|
our $VERSION = '0.01_03'; |
31
|
|
|
|
|
|
|
|
32
|
|
|
|
|
|
|
|
33
|
|
|
|
|
|
|
#===============================================================================# |
34
|
|
|
|
|
|
|
|
35
|
|
|
|
|
|
|
sub new { |
36
|
|
|
|
|
|
|
my $proto = shift; |
37
|
|
|
|
|
|
|
|
38
|
|
|
|
|
|
|
my %opts = validate(@_, { |
39
|
|
|
|
|
|
|
dates => { type => ARRAYREF, optional => 1 }, |
40
|
|
|
|
|
|
|
profile => { type => SCALAR | OBJECT | HASHREF, optional => 1 }, |
41
|
|
|
|
|
|
|
#stdev_limit => { type => SCALAR, default => 2 }, |
42
|
|
|
|
|
|
|
}); |
43
|
|
|
|
|
|
|
|
44
|
|
|
|
|
|
|
my $class = ref( $proto ) || $proto; |
45
|
|
|
|
|
|
|
my $self = { #Will need to allow for params passed to constructor |
46
|
|
|
|
|
|
|
dates => [], |
47
|
|
|
|
|
|
|
distinct_buckets => {}, |
48
|
|
|
|
|
|
|
interval_buckets => {}, |
49
|
|
|
|
|
|
|
total_epoch_interval => 0, |
50
|
|
|
|
|
|
|
largest_epoch_interval => 0, |
51
|
|
|
|
|
|
|
smallest_epoch_interval => 0, |
52
|
|
|
|
|
|
|
mean_epoch_interval => 0, |
53
|
|
|
|
|
|
|
|
54
|
|
|
|
|
|
|
#Whether this data set has been trained or not |
55
|
|
|
|
|
|
|
trained => 0, |
56
|
|
|
|
|
|
|
}; |
57
|
|
|
|
|
|
|
bless($self, $class); |
58
|
|
|
|
|
|
|
|
59
|
|
|
|
|
|
|
$opts{profile} = 'default' if ! $opts{profile}; |
60
|
|
|
|
|
|
|
|
61
|
|
|
|
|
|
|
$self->profile( $opts{profile} ); |
62
|
|
|
|
|
|
|
|
63
|
|
|
|
|
|
|
return $self; |
64
|
|
|
|
|
|
|
} |
65
|
|
|
|
|
|
|
|
66
|
|
|
|
|
|
|
# Get or set list of dates |
67
|
|
|
|
|
|
|
# ***NOTE: Should make this validate for 'can' on the DateTime methods we need and on 'isa' for DateTime |
68
|
|
|
|
|
|
|
sub dates { |
69
|
|
|
|
|
|
|
my $self = shift; |
70
|
|
|
|
|
|
|
my ($dates) = @_; |
71
|
|
|
|
|
|
|
|
72
|
|
|
|
|
|
|
validate_pos(@_, { type => ARRAYREF, optional => 1 }); |
73
|
|
|
|
|
|
|
|
74
|
|
|
|
|
|
|
if (! defined $dates) { |
75
|
|
|
|
|
|
|
return wantarray ? @{$self->{dates}} : $self->{dates}; |
76
|
|
|
|
|
|
|
} |
77
|
|
|
|
|
|
|
elsif (defined $dates) { |
78
|
|
|
|
|
|
|
foreach my $date (@$dates) { |
79
|
|
|
|
|
|
|
$self->_trim_date( $date ); |
80
|
|
|
|
|
|
|
$self->add_date($date); |
81
|
|
|
|
|
|
|
} |
82
|
|
|
|
|
|
|
} |
83
|
|
|
|
|
|
|
|
84
|
|
|
|
|
|
|
return 1; |
85
|
|
|
|
|
|
|
} |
86
|
|
|
|
|
|
|
|
87
|
|
|
|
|
|
|
# Add a date to the list of dates |
88
|
|
|
|
|
|
|
sub add_date { |
89
|
|
|
|
|
|
|
my $self = shift; |
90
|
|
|
|
|
|
|
my ($date) = @_; |
91
|
|
|
|
|
|
|
|
92
|
|
|
|
|
|
|
validate_pos(@_, { isa => 'DateTime' }); #***Or we could attempt to parse the date, or use can( epoch() ); |
93
|
|
|
|
|
|
|
|
94
|
|
|
|
|
|
|
$self->_trim_date( $date ); |
95
|
|
|
|
|
|
|
|
96
|
|
|
|
|
|
|
push(@{ $self->{dates} }, $date); |
97
|
|
|
|
|
|
|
|
98
|
|
|
|
|
|
|
return 1; |
99
|
|
|
|
|
|
|
} |
100
|
|
|
|
|
|
|
|
101
|
|
|
|
|
|
|
#Get or set the profile for this predictor |
102
|
|
|
|
|
|
|
sub profile { |
103
|
|
|
|
|
|
|
my $self = shift; |
104
|
|
|
|
|
|
|
my ($profile) = @_; # $profile can be a string specifying a profile name that is provided by default, or a profile object, or options to create a new profile |
105
|
|
|
|
|
|
|
|
106
|
|
|
|
|
|
|
validate_pos(@_, { type => SCALAR | OBJECT | HASHREF, optional => 1 }); |
107
|
|
|
|
|
|
|
|
108
|
|
|
|
|
|
|
# If no profile is provided, return the current profile |
109
|
|
|
|
|
|
|
if (! defined $profile || ! $profile) { return $self->{profile}; } |
110
|
|
|
|
|
|
|
|
111
|
|
|
|
|
|
|
my $new_profile; |
112
|
|
|
|
|
|
|
|
113
|
|
|
|
|
|
|
# Profile is an actual DTP::Profile object |
114
|
|
|
|
|
|
|
if (Scalar::Util::blessed($profile) && $profile->can('buckets')) { |
115
|
|
|
|
|
|
|
$new_profile = $profile; |
116
|
|
|
|
|
|
|
} |
117
|
|
|
|
|
|
|
# Profile is a hashref of options to create a new DTP::Profile object with |
118
|
|
|
|
|
|
|
elsif (ref($profile) eq 'HASH') { |
119
|
|
|
|
|
|
|
$new_profile = DateTime::Event::Predict::Profile->new( |
120
|
|
|
|
|
|
|
%$profile, |
121
|
|
|
|
|
|
|
); |
122
|
|
|
|
|
|
|
} |
123
|
|
|
|
|
|
|
# Profile is the name of a profile alias |
124
|
|
|
|
|
|
|
else { |
125
|
|
|
|
|
|
|
$new_profile = DateTime::Event::Predict::Profile->new( profile => $profile ); |
126
|
|
|
|
|
|
|
} |
127
|
|
|
|
|
|
|
|
128
|
|
|
|
|
|
|
# Add the distinct buckets |
129
|
|
|
|
|
|
|
foreach my $bucket ( $new_profile->_distinct_buckets() ) { |
130
|
|
|
|
|
|
|
$self->{distinct_buckets}->{ $bucket->name } = { |
131
|
|
|
|
|
|
|
accessor => $bucket->accessor, |
132
|
|
|
|
|
|
|
duration => $bucket->duration, |
133
|
|
|
|
|
|
|
order => $bucket->order, |
134
|
|
|
|
|
|
|
weight => $bucket->weight, |
135
|
|
|
|
|
|
|
buckets => {}, |
136
|
|
|
|
|
|
|
}; |
137
|
|
|
|
|
|
|
} |
138
|
|
|
|
|
|
|
|
139
|
|
|
|
|
|
|
# Add the interval buckets |
140
|
|
|
|
|
|
|
foreach my $bucket ( $new_profile->_interval_buckets() ) { |
141
|
|
|
|
|
|
|
$self->{interval_buckets}->{ $bucket->name } = { |
142
|
|
|
|
|
|
|
accessor => $bucket->accessor, |
143
|
|
|
|
|
|
|
order => $bucket->order, |
144
|
|
|
|
|
|
|
weight => $bucket->weight, |
145
|
|
|
|
|
|
|
buckets => {}, |
146
|
|
|
|
|
|
|
}; |
147
|
|
|
|
|
|
|
} |
148
|
|
|
|
|
|
|
|
149
|
|
|
|
|
|
|
$self->{profile} = $new_profile; |
150
|
|
|
|
|
|
|
|
151
|
|
|
|
|
|
|
return 1; |
152
|
|
|
|
|
|
|
} |
153
|
|
|
|
|
|
|
|
154
|
|
|
|
|
|
|
# Gather statistics about the dates |
155
|
|
|
|
|
|
|
sub train { |
156
|
|
|
|
|
|
|
my $self = shift; |
157
|
|
|
|
|
|
|
|
158
|
|
|
|
|
|
|
# Sort the dates chronologically |
159
|
|
|
|
|
|
|
my @dates = sort { $a->hires_epoch() <=> $b->hires_epoch() } @{ $self->{dates} }; #*** Need to convert this to DateTime->compare($dt1, $dt2) |
160
|
|
|
|
|
|
|
|
161
|
|
|
|
|
|
|
# Last and first dates |
162
|
|
|
|
|
|
|
$self->{last_date} = $dates[$#dates]; |
163
|
|
|
|
|
|
|
$self->{first_date} = $dates[0]; |
164
|
|
|
|
|
|
|
|
165
|
|
|
|
|
|
|
# Clear out anything already in the the buckets |
166
|
|
|
|
|
|
|
foreach my $bucket (values %{$self->{distinct_buckets}}, values %{$self->{interval_buckets}} ) { |
167
|
|
|
|
|
|
|
$bucket->{buckets} = {}; |
168
|
|
|
|
|
|
|
} |
169
|
|
|
|
|
|
|
|
170
|
|
|
|
|
|
|
my $prev_date; |
171
|
|
|
|
|
|
|
foreach my $index (0 .. $#{ $self->{dates} }) { |
172
|
|
|
|
|
|
|
# The date to work on |
173
|
|
|
|
|
|
|
my $date = $dates[ $index ]; |
174
|
|
|
|
|
|
|
|
175
|
|
|
|
|
|
|
# Get which dates were before and after the date we're working on |
176
|
|
|
|
|
|
|
my ($before, $after); |
177
|
|
|
|
|
|
|
if ($index > 0) { $before = $dates[ $index - 1 ]; } |
178
|
|
|
|
|
|
|
if ($index < $#{ $self->{dates} }) { $after = $dates[ $index + 1 ]; } |
179
|
|
|
|
|
|
|
|
180
|
|
|
|
|
|
|
# Increment the date-part buckets |
181
|
|
|
|
|
|
|
while (my ($name, $dbucket) = each %{ $self->{distinct_buckets} }) { |
182
|
|
|
|
|
|
|
# Get the accessor method by using can() |
183
|
|
|
|
|
|
|
my $cref = $date->can( $dbucket->{accessor} ); |
184
|
|
|
|
|
|
|
croak "Can't call accessor '" . $dbucket->{accessor} . "' on " . ref($date) . " object" unless $cref; |
185
|
|
|
|
|
|
|
|
186
|
|
|
|
|
|
|
# Increment the number of instances for the value given when we use this bucket's accessor on $date |
187
|
|
|
|
|
|
|
$dbucket->{buckets}->{ &$cref($date) }++; |
188
|
|
|
|
|
|
|
} |
189
|
|
|
|
|
|
|
|
190
|
|
|
|
|
|
|
# If this is the first date we have nothing to diff, so we'll skip on to the next one |
191
|
|
|
|
|
|
|
if (! $prev_date) { $prev_date = $date; next; } |
192
|
|
|
|
|
|
|
|
193
|
|
|
|
|
|
|
# Get a DateTime::Duration object representing the diff between the dates |
194
|
|
|
|
|
|
|
my $dur = $date->subtract_datetime( $prev_date ); |
195
|
|
|
|
|
|
|
|
196
|
|
|
|
|
|
|
# Increment the interval buckets |
197
|
|
|
|
|
|
|
# Intervals: here we default to the largest interval that we can see. So, for instance, if |
198
|
|
|
|
|
|
|
# there is a difference of months we will not increment anything smaller than that. |
199
|
|
|
|
|
|
|
while (my ($name, $bucket) = each %{ $self->{interval_buckets} }) { |
200
|
|
|
|
|
|
|
my $cref = $dur->can( $bucket->{accessor} ); |
201
|
|
|
|
|
|
|
croak "Can't call accessor '" . $bucket->{accessor} . "' on " . ref($dur) . " object" unless $cref; |
202
|
|
|
|
|
|
|
my $interval = &$cref($dur); |
203
|
|
|
|
|
|
|
$bucket->{buckets}->{ $interval }++; |
204
|
|
|
|
|
|
|
} |
205
|
|
|
|
|
|
|
|
206
|
|
|
|
|
|
|
# Add the difference between dates in epoch seconds |
207
|
|
|
|
|
|
|
my $epoch_interval = $date->hires_epoch() - $prev_date->hires_epoch(); |
208
|
|
|
|
|
|
|
|
209
|
|
|
|
|
|
|
### Epoch interval: $epoch_interval |
210
|
|
|
|
|
|
|
|
211
|
|
|
|
|
|
|
$self->{total_epoch_interval} += $epoch_interval; |
212
|
|
|
|
|
|
|
|
213
|
|
|
|
|
|
|
# Set the current date to this date |
214
|
|
|
|
|
|
|
$prev_date = $date; |
215
|
|
|
|
|
|
|
} |
216
|
|
|
|
|
|
|
|
217
|
|
|
|
|
|
|
# Average interval between dates in epoch seconds |
218
|
|
|
|
|
|
|
$self->{mean_epoch_interval} = $self->{total_epoch_interval} / (scalar @dates - 1); #Divide total interval by number of intervals |
219
|
|
|
|
|
|
|
|
220
|
|
|
|
|
|
|
# Mark this object as being trained |
221
|
|
|
|
|
|
|
$self->{trained}++; |
222
|
|
|
|
|
|
|
} |
223
|
|
|
|
|
|
|
|
224
|
|
|
|
|
|
|
sub predict { |
225
|
|
|
|
|
|
|
my $self = shift; |
226
|
|
|
|
|
|
|
|
227
|
|
|
|
|
|
|
my %opts = validate(@_, { |
228
|
|
|
|
|
|
|
max_predictions => { type => SCALAR, optional => 1 }, # How many predictions to return |
229
|
|
|
|
|
|
|
stdev_limit => { type => SCALAR, default => 2 }, # Number of standard deviations to search through, default to 2 |
230
|
|
|
|
|
|
|
min_date => { isa => 'DateTime', optional => 1 }, # If set, make no prediction before 'min_date' |
231
|
|
|
|
|
|
|
callbacks => { type => ARRAYREF, optional => 1 }, # Arrayref of coderefs to call when making predictions |
232
|
|
|
|
|
|
|
}); |
233
|
|
|
|
|
|
|
|
234
|
|
|
|
|
|
|
# Force max predictions to one if we were called in scalar context |
235
|
|
|
|
|
|
|
if (! defined $opts{'max_predictions'}) { |
236
|
|
|
|
|
|
|
$opts{'max_predictions'} = 1 if ! wantarray; |
237
|
|
|
|
|
|
|
} |
238
|
|
|
|
|
|
|
|
239
|
|
|
|
|
|
|
# Train this set of dates if they're not already trained |
240
|
|
|
|
|
|
|
$self->train if ! $self->_is_trained; |
241
|
|
|
|
|
|
|
|
242
|
|
|
|
|
|
|
# Make a copy of the distinct and interval bucket hashes so we can mess with them |
243
|
|
|
|
|
|
|
my %distinct_buckets = %{ $self->{distinct_buckets} }; |
244
|
|
|
|
|
|
|
my %interval_buckets = %{ $self->{interval_buckets} }; |
245
|
|
|
|
|
|
|
|
246
|
|
|
|
|
|
|
# Figure the mean, variance, and standard deviation for each bucket |
247
|
|
|
|
|
|
|
foreach my $bucket (values %distinct_buckets, values %interval_buckets) { |
248
|
|
|
|
|
|
|
my ($mean, $variance, $stdev) = $self->_bucket_statistics($bucket); |
249
|
|
|
|
|
|
|
|
250
|
|
|
|
|
|
|
$bucket->{mean} = $mean; |
251
|
|
|
|
|
|
|
$bucket->{variance} = $variance; |
252
|
|
|
|
|
|
|
$bucket->{stdev} = $stdev; |
253
|
|
|
|
|
|
|
} |
254
|
|
|
|
|
|
|
|
255
|
|
|
|
|
|
|
# Get the most recent of the provided dates by sorting them by their epoch seconds |
256
|
|
|
|
|
|
|
my $most_recent_date = (sort { $b->hires_epoch() <=> $a->hires_epoch() } @{ $self->{dates} })[0]; |
257
|
|
|
|
|
|
|
|
258
|
|
|
|
|
|
|
# Make a starting search date that has been moved ahead by the average interval beteween dates (in epoch seconds) |
259
|
|
|
|
|
|
|
my $duration = new DateTime::Duration( |
260
|
|
|
|
|
|
|
seconds => $self->{mean_epoch_interval}, # **Might need to round off hires second info here? |
261
|
|
|
|
|
|
|
); |
262
|
|
|
|
|
|
|
my $start_date = $most_recent_date + $duration; |
263
|
|
|
|
|
|
|
|
264
|
|
|
|
|
|
|
# A hash of predictions, dates are keyed by their hires_epoch() value |
265
|
|
|
|
|
|
|
my %predictions = (); |
266
|
|
|
|
|
|
|
|
267
|
|
|
|
|
|
|
# Start with using the distinct buckets to make predictions |
268
|
|
|
|
|
|
|
if (%distinct_buckets) { |
269
|
|
|
|
|
|
|
# Get a list of buckets after sorting the buckets from largest date part to smallest (i.e. year->month->day->hour ... microsecond, etc) |
270
|
|
|
|
|
|
|
my @distinct_bucket_keys = sort { $self->{distinct_buckets}->{ $b }->{order} <=> $self->{distinct_buckets}->{ $a }->{order} } keys %distinct_buckets; |
271
|
|
|
|
|
|
|
|
272
|
|
|
|
|
|
|
# Get the first bucket name |
273
|
|
|
|
|
|
|
my $first_bucket_name = shift @distinct_bucket_keys; |
274
|
|
|
|
|
|
|
|
275
|
|
|
|
|
|
|
# Start recursively descending down into the various date parts, searching in each one |
276
|
|
|
|
|
|
|
$self->_date_descend_distinct( |
277
|
|
|
|
|
|
|
%opts, |
278
|
|
|
|
|
|
|
|
279
|
|
|
|
|
|
|
date => $start_date, |
280
|
|
|
|
|
|
|
most_recent_date => $most_recent_date, |
281
|
|
|
|
|
|
|
bucket_name => $first_bucket_name, |
282
|
|
|
|
|
|
|
distinct_buckets => \%distinct_buckets, |
283
|
|
|
|
|
|
|
distinct_bucket_keys => \@distinct_bucket_keys, |
284
|
|
|
|
|
|
|
predictions => \%predictions, |
285
|
|
|
|
|
|
|
); |
286
|
|
|
|
|
|
|
|
287
|
|
|
|
|
|
|
# Now that we (hopefully) have some predictions, put them each through _interval_check to check |
288
|
|
|
|
|
|
|
# the predictiosn against the interval bucket statistics |
289
|
|
|
|
|
|
|
if (%interval_buckets) { |
290
|
|
|
|
|
|
|
while (my ($hires, $prediction) = each %predictions) { |
291
|
|
|
|
|
|
|
# Delete the date from the predictions hash if it's not good according to the interval statistics |
292
|
|
|
|
|
|
|
if (! $self->_interval_check( $prediction )) { |
293
|
|
|
|
|
|
|
delete $predictions{ $hires }; |
294
|
|
|
|
|
|
|
} |
295
|
|
|
|
|
|
|
} |
296
|
|
|
|
|
|
|
} |
297
|
|
|
|
|
|
|
} |
298
|
|
|
|
|
|
|
# No distinct buckets, just interval buckets |
299
|
|
|
|
|
|
|
elsif (%interval_buckets) { |
300
|
|
|
|
|
|
|
# Get a list of buckets after sorting the buckets from largest interval to smallest (i.e. years->months->days->hours, etc) |
301
|
|
|
|
|
|
|
my @interval_bucket_keys = sort { $self->{interval_buckets}->{ $b }->{order} <=> $self->{interval_buckets}->{ $a }->{order} } keys %interval_buckets; |
302
|
|
|
|
|
|
|
|
303
|
|
|
|
|
|
|
# Get the first bucket name |
304
|
|
|
|
|
|
|
my $first_bucket_name = shift @interval_bucket_keys; |
305
|
|
|
|
|
|
|
|
306
|
|
|
|
|
|
|
# Start recursively descending down into the date interval types, searching in each one |
307
|
|
|
|
|
|
|
$self->_date_descend_interval( |
308
|
|
|
|
|
|
|
%opts, |
309
|
|
|
|
|
|
|
|
310
|
|
|
|
|
|
|
date => $start_date, |
311
|
|
|
|
|
|
|
most_recent_date => $most_recent_date, |
312
|
|
|
|
|
|
|
bucket_name => $first_bucket_name, |
313
|
|
|
|
|
|
|
interval_buckets => \%interval_buckets, |
314
|
|
|
|
|
|
|
interval_bucket_keys => \@interval_bucket_keys, |
315
|
|
|
|
|
|
|
predictions => \%predictions, |
316
|
|
|
|
|
|
|
); |
317
|
|
|
|
|
|
|
} |
318
|
|
|
|
|
|
|
# WTF, no buckets. That's bad! |
319
|
|
|
|
|
|
|
else { |
320
|
|
|
|
|
|
|
croak("No buckets supplied!"); |
321
|
|
|
|
|
|
|
} |
322
|
|
|
|
|
|
|
|
323
|
|
|
|
|
|
|
# Sort the predictions by their total deviation |
324
|
|
|
|
|
|
|
my @predictions = sort { $a->{_dtp_deviation} <=> $b->{_dtp_deviation} } values %predictions; |
325
|
|
|
|
|
|
|
|
326
|
|
|
|
|
|
|
return wantarray ? @predictions : $predictions[0]; |
327
|
|
|
|
|
|
|
} |
328
|
|
|
|
|
|
|
|
329
|
|
|
|
|
|
|
# Descend down into the distinct date parts, looking for predictions |
330
|
|
|
|
|
|
|
sub _date_descend_distinct { |
331
|
|
|
|
|
|
|
my $self = shift; |
332
|
|
|
|
|
|
|
#my %opts = @_; |
333
|
|
|
|
|
|
|
|
334
|
|
|
|
|
|
|
# Validate the options |
335
|
|
|
|
|
|
|
my %opts = validate(@_, { |
336
|
|
|
|
|
|
|
date => { isa => 'DateTime' }, # The date to start searching in |
337
|
|
|
|
|
|
|
most_recent_date => { isa => 'DateTime' }, # The most recent date of the dates provided |
338
|
|
|
|
|
|
|
bucket_name => { type => SCALAR }, # The bucket (date-part) to start searching in |
339
|
|
|
|
|
|
|
distinct_buckets => { type => HASHREF }, # A hashref of all buckets to use when looking for good predictions |
340
|
|
|
|
|
|
|
distinct_bucket_keys => { type => ARRAYREF }, # A list of bucket names that we shift out of to get the next bucket to use |
341
|
|
|
|
|
|
|
stdev_limit => { type => SCALAR }, # The limit of how many standard deviations to search through |
342
|
|
|
|
|
|
|
predictions => { type => HASHREF }, # A hashref of predictions we find |
343
|
|
|
|
|
|
|
max_predictions => { type => SCALAR, optional => 1 }, # The maxmimum number of predictions to return (prevents overly long searches) |
344
|
|
|
|
|
|
|
min_date => { isa => 'DateTime', optional => 1 }, # If set, make no prediction before 'min_date' |
345
|
|
|
|
|
|
|
callbacks => { type => ARRAYREF, optional => 1 }, # A list of custom coderefs that are called on each possible prediction |
346
|
|
|
|
|
|
|
}); |
347
|
|
|
|
|
|
|
|
348
|
|
|
|
|
|
|
# Copy the options over into simple scalars so it's easier on my eyes |
349
|
|
|
|
|
|
|
my $date = delete $opts{'date'}; # Delete these ones out as we'll be overwriting them below |
350
|
|
|
|
|
|
|
my $bucket_name = delete $opts{'bucket_name'}; |
351
|
|
|
|
|
|
|
my $distinct_buckets = $opts{'distinct_buckets'}; |
352
|
|
|
|
|
|
|
my $distinct_bucket_keys = $opts{'distinct_bucket_keys'}; |
353
|
|
|
|
|
|
|
my $stdev_limit = $opts{'stdev_limit'}; |
354
|
|
|
|
|
|
|
my $predictions = $opts{'predictions'}; |
355
|
|
|
|
|
|
|
my $max_predictions = $opts{'max_predictions'}; |
356
|
|
|
|
|
|
|
my $callbacks = $opts{'callbacks'}; |
357
|
|
|
|
|
|
|
|
358
|
|
|
|
|
|
|
# We've reached our max number of predictions, return |
359
|
|
|
|
|
|
|
return 1 if defined $max_predictions && (scalar keys %$predictions) >= $max_predictions; |
360
|
|
|
|
|
|
|
|
361
|
|
|
|
|
|
|
# Get the actual bucket hash for this bucket name |
362
|
|
|
|
|
|
|
my $bucket = $distinct_buckets->{ $bucket_name }; |
363
|
|
|
|
|
|
|
|
364
|
|
|
|
|
|
|
# The search range is the standard deviation multiplied by the number of standard deviations to search through |
365
|
|
|
|
|
|
|
my $search_range = ceil( $bucket->{stdev} * $stdev_limit ); |
366
|
|
|
|
|
|
|
|
367
|
|
|
|
|
|
|
#The next bucket to search down into |
368
|
|
|
|
|
|
|
my $next_bucket_name = ""; |
369
|
|
|
|
|
|
|
if (scalar @$distinct_bucket_keys > 0) { |
370
|
|
|
|
|
|
|
$next_bucket_name = shift @$distinct_bucket_keys; |
371
|
|
|
|
|
|
|
} |
372
|
|
|
|
|
|
|
|
373
|
|
|
|
|
|
|
foreach my $search_inc ( 0 .. $search_range ) { |
374
|
|
|
|
|
|
|
# Make an inverted search increment so we can search backwards |
375
|
|
|
|
|
|
|
my $neg_search_inc = $search_inc * -1; |
376
|
|
|
|
|
|
|
|
377
|
|
|
|
|
|
|
# Put forwards and backwards in the searches |
378
|
|
|
|
|
|
|
my @searches = ($search_inc, $neg_search_inc); |
379
|
|
|
|
|
|
|
|
380
|
|
|
|
|
|
|
# Make sure we only search on 0 once (i.e. 0 * -1 == 0) |
381
|
|
|
|
|
|
|
@searches = (0) if $search_inc == 0; |
382
|
|
|
|
|
|
|
|
383
|
|
|
|
|
|
|
foreach my $increment (@searches) { |
384
|
|
|
|
|
|
|
# We've reached our max number of predictions, return |
385
|
|
|
|
|
|
|
return 1 if defined $max_predictions && (scalar keys %$predictions) >= $max_predictions; |
386
|
|
|
|
|
|
|
|
387
|
|
|
|
|
|
|
# Make a duration object using the accessor for this bucket |
388
|
|
|
|
|
|
|
my $duration_increment = new DateTime::Duration( $bucket->{duration} => $increment ); |
389
|
|
|
|
|
|
|
|
390
|
|
|
|
|
|
|
# Get the new date |
391
|
|
|
|
|
|
|
my $new_date = $date + $duration_increment; |
392
|
|
|
|
|
|
|
|
393
|
|
|
|
|
|
|
# Trim the date down to just the date parts we care about |
394
|
|
|
|
|
|
|
$self->_trim_date( $new_date ); |
395
|
|
|
|
|
|
|
|
396
|
|
|
|
|
|
|
# Skip this date if it's before or on the most recent date |
397
|
|
|
|
|
|
|
if (DateTime->compare( $new_date, $opts{'most_recent_date'} ) <= 0) { # New date is before the most recent one, or is same as most recent one |
398
|
|
|
|
|
|
|
next; |
399
|
|
|
|
|
|
|
} |
400
|
|
|
|
|
|
|
|
401
|
|
|
|
|
|
|
# Skip this date if the "min_date" option is set, and it's before or on that date |
402
|
|
|
|
|
|
|
if ($opts{'min_date'} && DateTime->compare($new_date, $opts{'min_date'}) <= 0) { |
403
|
|
|
|
|
|
|
next; |
404
|
|
|
|
|
|
|
} |
405
|
|
|
|
|
|
|
|
406
|
|
|
|
|
|
|
# If we have no more buckets to search into, determine if this date is a good prediction |
407
|
|
|
|
|
|
|
if (! $next_bucket_name) { |
408
|
|
|
|
|
|
|
if ($self->_distinct_check( %opts, date => $new_date )) { |
409
|
|
|
|
|
|
|
$predictions->{ $new_date->hires_epoch() } = $new_date; |
410
|
|
|
|
|
|
|
} |
411
|
|
|
|
|
|
|
} |
412
|
|
|
|
|
|
|
#If we're not at the smallest bucket, keep searching! |
413
|
|
|
|
|
|
|
else { |
414
|
|
|
|
|
|
|
$self->_date_descend_distinct( |
415
|
|
|
|
|
|
|
%opts, |
416
|
|
|
|
|
|
|
date => $new_date, |
417
|
|
|
|
|
|
|
bucket_name => $next_bucket_name, |
418
|
|
|
|
|
|
|
); |
419
|
|
|
|
|
|
|
} |
420
|
|
|
|
|
|
|
} |
421
|
|
|
|
|
|
|
} |
422
|
|
|
|
|
|
|
|
423
|
|
|
|
|
|
|
return 1; |
424
|
|
|
|
|
|
|
} |
425
|
|
|
|
|
|
|
|
426
|
|
|
|
|
|
|
# Descend down into the date intervals, looking for predictions |
427
|
|
|
|
|
|
|
sub _date_descend_interval { |
428
|
|
|
|
|
|
|
my $self = shift; |
429
|
|
|
|
|
|
|
|
430
|
|
|
|
|
|
|
# Validate the options |
431
|
|
|
|
|
|
|
my %opts = validate(@_, { |
432
|
|
|
|
|
|
|
date => { isa => 'DateTime' }, # The date to start searching in |
433
|
|
|
|
|
|
|
most_recent_date => { isa => 'DateTime' }, # The most recent date of the dates provided |
434
|
|
|
|
|
|
|
bucket_name => { type => SCALAR }, # The bucket (date-part) to start searching in |
435
|
|
|
|
|
|
|
interval_buckets => { type => HASHREF }, # A hashref of all buckets to use when looking for good predictions |
436
|
|
|
|
|
|
|
interval_bucket_keys => { type => ARRAYREF }, # A list of bucket names that we shift out of to get the next bucket to use |
437
|
|
|
|
|
|
|
stdev_limit => { type => SCALAR }, # The limit of how many standard deviations to search through |
438
|
|
|
|
|
|
|
predictions => { type => HASHREF }, # A hashref of predictions we find |
439
|
|
|
|
|
|
|
max_predictions => { type => SCALAR, optional => 1 }, # The maxmimum number of predictions to return (prevents overly long searches) |
440
|
|
|
|
|
|
|
min_date => { isa => 'DateTime', optional => 1 }, # If set, make no prediction before 'min_date' |
441
|
|
|
|
|
|
|
callbacks => { type => ARRAYREF, optional => 1 }, # A list of custom coderefs that are called on each possible prediction |
442
|
|
|
|
|
|
|
}); |
443
|
|
|
|
|
|
|
|
444
|
|
|
|
|
|
|
# Copy the options over into simple scalars so it's easier on my eyes |
445
|
|
|
|
|
|
|
my $date = delete $opts{'date'}; # Delete these ones out as we'll be overwriting them below |
446
|
|
|
|
|
|
|
my $bucket_name = delete $opts{'bucket_name'}; |
447
|
|
|
|
|
|
|
my $interval_buckets = $opts{'interval_buckets'}; |
448
|
|
|
|
|
|
|
my $interval_bucket_keys = $opts{'interval_bucket_keys'}; |
449
|
|
|
|
|
|
|
my $stdev_limit = $opts{'stdev_limit'}; |
450
|
|
|
|
|
|
|
my $predictions = $opts{'predictions'}; |
451
|
|
|
|
|
|
|
my $max_predictions = $opts{'max_predictions'}; |
452
|
|
|
|
|
|
|
my $callbacks = $opts{'callbacks'}; |
453
|
|
|
|
|
|
|
|
454
|
|
|
|
|
|
|
# We've reached our max number of predictions, return |
455
|
|
|
|
|
|
|
return 1 if defined $max_predictions && (scalar keys %$predictions) >= $max_predictions; |
456
|
|
|
|
|
|
|
|
457
|
|
|
|
|
|
|
# Get the actual bucket hash for this bucket name |
458
|
|
|
|
|
|
|
my $bucket = $interval_buckets->{ $bucket_name }; |
459
|
|
|
|
|
|
|
|
460
|
|
|
|
|
|
|
# The search range is the standard deviation multiplied by the number of standard deviations to search through |
461
|
|
|
|
|
|
|
my $search_range = ceil( $bucket->{stdev} * $stdev_limit ); |
462
|
|
|
|
|
|
|
|
463
|
|
|
|
|
|
|
#The next bucket to search down into |
464
|
|
|
|
|
|
|
my $next_bucket_name = ""; |
465
|
|
|
|
|
|
|
if (scalar @$interval_bucket_keys > 0) { |
466
|
|
|
|
|
|
|
$next_bucket_name = shift @$interval_bucket_keys; |
467
|
|
|
|
|
|
|
} |
468
|
|
|
|
|
|
|
|
469
|
|
|
|
|
|
|
foreach my $search_inc ( 0 .. $search_range ) { |
470
|
|
|
|
|
|
|
# Make an inverted search increment so we can search backwards |
471
|
|
|
|
|
|
|
my $neg_search_inc = $search_inc * -1; |
472
|
|
|
|
|
|
|
|
473
|
|
|
|
|
|
|
# Put forwards and backwards in the searches |
474
|
|
|
|
|
|
|
my @searches = ($search_inc, $neg_search_inc); |
475
|
|
|
|
|
|
|
|
476
|
|
|
|
|
|
|
# Make sure we only search on 0 once (i.e. 0 * -1 == 0) |
477
|
|
|
|
|
|
|
@searches = (0) if $search_inc == 0; |
478
|
|
|
|
|
|
|
|
479
|
|
|
|
|
|
|
foreach my $increment (@searches) { |
480
|
|
|
|
|
|
|
# We've reached our max number of predictions, return |
481
|
|
|
|
|
|
|
return 1 if defined $max_predictions && (scalar keys %$predictions) >= $max_predictions; |
482
|
|
|
|
|
|
|
|
483
|
|
|
|
|
|
|
# Make a duration object using the accessor for this bucket |
484
|
|
|
|
|
|
|
my $duration_increment = new DateTime::Duration( $bucket->{accessor} => $increment ); |
485
|
|
|
|
|
|
|
|
486
|
|
|
|
|
|
|
# Get the new date |
487
|
|
|
|
|
|
|
my $new_date = $date + $duration_increment; |
488
|
|
|
|
|
|
|
|
489
|
|
|
|
|
|
|
# Trim the date down to just the date parts we care about |
490
|
|
|
|
|
|
|
$self->_trim_date( $new_date ); |
491
|
|
|
|
|
|
|
|
492
|
|
|
|
|
|
|
# Skip this date if it's before or on the most recent date |
493
|
|
|
|
|
|
|
if (DateTime->compare( $new_date, $opts{'most_recent_date'} ) <= 0) { # New date is before the most recent one, or is same as most recent one |
494
|
|
|
|
|
|
|
next; |
495
|
|
|
|
|
|
|
} |
496
|
|
|
|
|
|
|
|
497
|
|
|
|
|
|
|
# Skip this date if the "min_date" option is set, and it's before or on that date |
498
|
|
|
|
|
|
|
if ($opts{'min_date'} && DateTime->compare($new_date, $opts{'min_date'}) <= 0) { |
499
|
|
|
|
|
|
|
next; |
500
|
|
|
|
|
|
|
} |
501
|
|
|
|
|
|
|
|
502
|
|
|
|
|
|
|
# If we have no more buckets to search into, determine if this date is a good prediction |
503
|
|
|
|
|
|
|
if (! $next_bucket_name) { |
504
|
|
|
|
|
|
|
if ($self->_interval_check( %opts, date => $new_date )) { |
505
|
|
|
|
|
|
|
$predictions->{ $new_date->hires_epoch() } = $new_date; |
506
|
|
|
|
|
|
|
} |
507
|
|
|
|
|
|
|
} |
508
|
|
|
|
|
|
|
#If we're not at the smallest bucket, keep searching! |
509
|
|
|
|
|
|
|
else { |
510
|
|
|
|
|
|
|
$self->_date_descend_interval( |
511
|
|
|
|
|
|
|
%opts, |
512
|
|
|
|
|
|
|
date => $new_date, |
513
|
|
|
|
|
|
|
bucket_name => $next_bucket_name, |
514
|
|
|
|
|
|
|
); |
515
|
|
|
|
|
|
|
} |
516
|
|
|
|
|
|
|
} |
517
|
|
|
|
|
|
|
} |
518
|
|
|
|
|
|
|
|
519
|
|
|
|
|
|
|
return 1; |
520
|
|
|
|
|
|
|
} |
521
|
|
|
|
|
|
|
|
522
|
|
|
|
|
|
|
# Check to see if a given date is good according to the supplied distinct buckets by going through each bucket |
523
|
|
|
|
|
|
|
# and comparing this date's deviation from that bucket's mean. If it is within the standard deviation for |
524
|
|
|
|
|
|
|
# each bucket then consider it a good match. |
525
|
|
|
|
|
|
|
sub _distinct_check { |
526
|
|
|
|
|
|
|
my $self = shift; |
527
|
|
|
|
|
|
|
|
528
|
|
|
|
|
|
|
# Temporarily allow extra options |
529
|
|
|
|
|
|
|
validation_options( allow_extra => 1 ); |
530
|
|
|
|
|
|
|
my %opts = validate(@_, { |
531
|
|
|
|
|
|
|
date => { isa => 'DateTime' }, # The date to check |
532
|
|
|
|
|
|
|
distinct_buckets => { type => HASHREF }, # List of enabled buckets |
533
|
|
|
|
|
|
|
callbacks => { type => ARRAYREF, optional => 1 }, # A list of custom coderefs that are called on each possible prediction |
534
|
|
|
|
|
|
|
}); |
535
|
|
|
|
|
|
|
validation_options( allow_extra => 0 ); |
536
|
|
|
|
|
|
|
|
537
|
|
|
|
|
|
|
my $date = $opts{'date'}; |
538
|
|
|
|
|
|
|
my $distinct_buckets = $opts{'distinct_buckets'}; |
539
|
|
|
|
|
|
|
my $callbacks = $opts{'callbacks'}; |
540
|
|
|
|
|
|
|
|
541
|
|
|
|
|
|
|
my $good = 1; |
542
|
|
|
|
|
|
|
my $date_deviation = 0; |
543
|
|
|
|
|
|
|
foreach my $bucket (values %$distinct_buckets) { |
544
|
|
|
|
|
|
|
# Get the value for this bucket's access for the $new_date |
545
|
|
|
|
|
|
|
my $cref = $date->can( $bucket->{accessor} ); |
546
|
|
|
|
|
|
|
my $datepart_val = &$cref($date); |
547
|
|
|
|
|
|
|
|
548
|
|
|
|
|
|
|
# If the deviation of this datepart from the mean is within the standard deviation, |
549
|
|
|
|
|
|
|
# this date ain't good. |
550
|
|
|
|
|
|
|
|
551
|
|
|
|
|
|
|
my $deviation = abs($datepart_val - $bucket->{mean}); |
552
|
|
|
|
|
|
|
$date_deviation += $deviation; |
553
|
|
|
|
|
|
|
|
554
|
|
|
|
|
|
|
if ($deviation > $bucket->{stdev} ) { |
555
|
|
|
|
|
|
|
$good = 0; |
556
|
|
|
|
|
|
|
last; |
557
|
|
|
|
|
|
|
} |
558
|
|
|
|
|
|
|
} |
559
|
|
|
|
|
|
|
|
560
|
|
|
|
|
|
|
# All the dateparts were within their standard deviations, check for callbacks and push this date into the set of predictions |
561
|
|
|
|
|
|
|
if ($good == 1) { |
562
|
|
|
|
|
|
|
# Stick the date's total deviation into the object so it can be used for sorting in predict() |
563
|
|
|
|
|
|
|
$date->{_dtp_deviation} += $date_deviation; |
564
|
|
|
|
|
|
|
|
565
|
|
|
|
|
|
|
# Run each hook we were passed |
566
|
|
|
|
|
|
|
foreach my $callback (@$callbacks) { |
567
|
|
|
|
|
|
|
# If any hook returns false, this date is a no-go and we can stop processing it |
568
|
|
|
|
|
|
|
if (! &$callback($date)) { |
569
|
|
|
|
|
|
|
$good = 0; |
570
|
|
|
|
|
|
|
last; |
571
|
|
|
|
|
|
|
} |
572
|
|
|
|
|
|
|
} |
573
|
|
|
|
|
|
|
|
574
|
|
|
|
|
|
|
# If the date is still considered good, return true |
575
|
|
|
|
|
|
|
if ($good == 1) { |
576
|
|
|
|
|
|
|
return 1; |
577
|
|
|
|
|
|
|
} |
578
|
|
|
|
|
|
|
# Otherwise return false |
579
|
|
|
|
|
|
|
else { |
580
|
|
|
|
|
|
|
return 0; |
581
|
|
|
|
|
|
|
} |
582
|
|
|
|
|
|
|
} |
583
|
|
|
|
|
|
|
} |
584
|
|
|
|
|
|
|
|
585
|
|
|
|
|
|
|
# Check to see if a given date is good according to the supplied interval buckets by going through each bucket |
586
|
|
|
|
|
|
|
# and comparing this date's deviation from that bucket's mean. If it is within the standard deviation for |
587
|
|
|
|
|
|
|
# each bucket then consider it a good match. |
588
|
|
|
|
|
|
|
sub _interval_check { |
589
|
|
|
|
|
|
|
my $self = shift; |
590
|
|
|
|
|
|
|
|
591
|
|
|
|
|
|
|
# Temporarily allow extra options |
592
|
|
|
|
|
|
|
validation_options( allow_extra => 1 ); |
593
|
|
|
|
|
|
|
my %opts = validate(@_, { |
594
|
|
|
|
|
|
|
date => { isa => 'DateTime' }, # The date prediction to check |
595
|
|
|
|
|
|
|
most_recent_date => { isa => 'DateTime' }, # The most recent date of the dates provided |
596
|
|
|
|
|
|
|
interval_buckets => { type => HASHREF }, # List of enabled interval buckets |
597
|
|
|
|
|
|
|
callbacks => { type => ARRAYREF, optional => 1 }, # A list of custom coderefs that are called on each possible prediction |
598
|
|
|
|
|
|
|
}); |
599
|
|
|
|
|
|
|
validation_options( allow_extra => 0 ); |
600
|
|
|
|
|
|
|
|
601
|
|
|
|
|
|
|
my $date = $opts{'date'}; |
602
|
|
|
|
|
|
|
my $most_recent_date = $opts{'most_recent_date'}; |
603
|
|
|
|
|
|
|
my $interval_buckets = $opts{'interval_buckets'}; |
604
|
|
|
|
|
|
|
my $callbacks = $opts{'callbacks'}; |
605
|
|
|
|
|
|
|
|
606
|
|
|
|
|
|
|
# Flag specifying whether the predicted date is "good" (within the standard deviation) or not |
607
|
|
|
|
|
|
|
my $good = 1; |
608
|
|
|
|
|
|
|
|
609
|
|
|
|
|
|
|
# Total deviation of the predicted date from each of the bucket standard deviations |
610
|
|
|
|
|
|
|
my $date_deviation = 0; |
611
|
|
|
|
|
|
|
|
612
|
|
|
|
|
|
|
# Get a duration object for the span between the most recent date supplied and the predicted date |
613
|
|
|
|
|
|
|
my $dur = $date->subtract_datetime( $most_recent_date ); |
614
|
|
|
|
|
|
|
|
615
|
|
|
|
|
|
|
foreach my $bucket (values %$interval_buckets) { |
616
|
|
|
|
|
|
|
my $cref = $dur->can( $bucket->{accessor} ); |
617
|
|
|
|
|
|
|
croak "Can't call accessor '" . $bucket->{accessor} . "' on " . ref($dur) . " object" unless $cref; |
618
|
|
|
|
|
|
|
my $interval = &$cref($dur); |
619
|
|
|
|
|
|
|
|
620
|
|
|
|
|
|
|
my $deviation = abs($interval - $bucket->{mean}); |
621
|
|
|
|
|
|
|
$date_deviation += $deviation; |
622
|
|
|
|
|
|
|
|
623
|
|
|
|
|
|
|
if ($deviation > $bucket->{stdev} ) { |
624
|
|
|
|
|
|
|
$good = 0; |
625
|
|
|
|
|
|
|
last; |
626
|
|
|
|
|
|
|
} |
627
|
|
|
|
|
|
|
} |
628
|
|
|
|
|
|
|
|
629
|
|
|
|
|
|
|
# All the dateparts were within their standard deviations, check for callbacks and push this date into the set of predictions |
630
|
|
|
|
|
|
|
if ($good == 1) { |
631
|
|
|
|
|
|
|
# Stick the date's total deviation into the object so it can be used for sorting in predict() |
632
|
|
|
|
|
|
|
$date->{_dtp_deviation} += $date_deviation; |
633
|
|
|
|
|
|
|
|
634
|
|
|
|
|
|
|
# Run each hook we were passed |
635
|
|
|
|
|
|
|
foreach my $callback (@$callbacks) { |
636
|
|
|
|
|
|
|
# If any hook returns false, this date is a no-go and we can stop processing it |
637
|
|
|
|
|
|
|
if (! &$callback($date)) { |
638
|
|
|
|
|
|
|
$good = 0; |
639
|
|
|
|
|
|
|
last; |
640
|
|
|
|
|
|
|
} |
641
|
|
|
|
|
|
|
} |
642
|
|
|
|
|
|
|
|
643
|
|
|
|
|
|
|
# If the date is still considered good, return true |
644
|
|
|
|
|
|
|
if ($good == 1) { |
645
|
|
|
|
|
|
|
return 1; |
646
|
|
|
|
|
|
|
} |
647
|
|
|
|
|
|
|
# Otherwise return false |
648
|
|
|
|
|
|
|
else { |
649
|
|
|
|
|
|
|
return 0; |
650
|
|
|
|
|
|
|
} |
651
|
|
|
|
|
|
|
} |
652
|
|
|
|
|
|
|
} |
653
|
|
|
|
|
|
|
|
654
|
|
|
|
|
|
|
# Get the mean, variance, and standard deviation for a bucket |
655
|
|
|
|
|
|
|
sub _bucket_statistics { |
656
|
|
|
|
|
|
|
my $self = shift; |
657
|
|
|
|
|
|
|
my $bucket = shift; |
658
|
|
|
|
|
|
|
|
659
|
|
|
|
|
|
|
my $total = 0; |
660
|
|
|
|
|
|
|
my $count = 0; |
661
|
|
|
|
|
|
|
while (my ($value, $occurances) = each %{ $bucket->{buckets} }) { |
662
|
|
|
|
|
|
|
# Gotta loop for each time the value has been found, incrementing the total by the value |
663
|
|
|
|
|
|
|
for (1 .. $occurances) { |
664
|
|
|
|
|
|
|
$total += $value; |
665
|
|
|
|
|
|
|
$count++; |
666
|
|
|
|
|
|
|
} |
667
|
|
|
|
|
|
|
} |
668
|
|
|
|
|
|
|
|
669
|
|
|
|
|
|
|
my $mean = $total / $count; |
670
|
|
|
|
|
|
|
|
671
|
|
|
|
|
|
|
# Get the variance |
672
|
|
|
|
|
|
|
my $total_variance = 0; |
673
|
|
|
|
|
|
|
while (my ($value, $occurances) = each %{ $bucket->{buckets} }) { |
674
|
|
|
|
|
|
|
# Gotta loop for each time the value has been found |
675
|
|
|
|
|
|
|
my $this_variance = ($value - $mean) ** 2; |
676
|
|
|
|
|
|
|
|
677
|
|
|
|
|
|
|
$total_variance += $this_variance * $occurances; |
678
|
|
|
|
|
|
|
} |
679
|
|
|
|
|
|
|
|
680
|
|
|
|
|
|
|
my $variance = $total_variance / $count; |
681
|
|
|
|
|
|
|
my $stdev = sqrt($variance); |
682
|
|
|
|
|
|
|
|
683
|
|
|
|
|
|
|
return ($mean, $variance, $stdev); |
684
|
|
|
|
|
|
|
} |
685
|
|
|
|
|
|
|
|
686
|
|
|
|
|
|
|
# Whether this instance has been trained by train() or not |
687
|
|
|
|
|
|
|
sub _is_trained { |
688
|
|
|
|
|
|
|
my $self = shift; |
689
|
|
|
|
|
|
|
|
690
|
|
|
|
|
|
|
return ($self->{trained} > 0) ? 1 : 0; |
691
|
|
|
|
|
|
|
} |
692
|
|
|
|
|
|
|
|
693
|
|
|
|
|
|
|
# Utility method to print out the dates added to this instance |
694
|
|
|
|
|
|
|
sub _print_dates { |
695
|
|
|
|
|
|
|
my $self = shift; |
696
|
|
|
|
|
|
|
|
697
|
|
|
|
|
|
|
foreach my $date (sort { $a->hires_epoch() <=> $b->hires_epoch() } @{ $self->{dates} }) { |
698
|
|
|
|
|
|
|
print $date->mdy('/') . ' ' . $date->hms . "\n"; |
699
|
|
|
|
|
|
|
} |
700
|
|
|
|
|
|
|
} |
701
|
|
|
|
|
|
|
|
702
|
|
|
|
|
|
|
# Trim the date parts that are smaller than the smallest one we care about. If we only care about |
703
|
|
|
|
|
|
|
# the year, month, and day, and during the initial search create an offset date that has an hour |
704
|
|
|
|
|
|
|
# or minute that is off from the most recent given date, then when we do a comparison to see if |
705
|
|
|
|
|
|
|
# we're predicting a date we've already been given it's possible that we could have that same |
706
|
|
|
|
|
|
|
# date, just with the hour and second set forward a bit. |
707
|
|
|
|
|
|
|
sub _trim_dates { |
708
|
|
|
|
|
|
|
my $self = shift; |
709
|
|
|
|
|
|
|
my (@dates) = @_; |
710
|
|
|
|
|
|
|
|
711
|
|
|
|
|
|
|
# Get the smallest bucket we have turned on |
712
|
|
|
|
|
|
|
my @buckets = (sort { $a->order <=> $b->order } grep { $_->on && $_->trimmable } $self->profile->buckets)[0]; |
713
|
|
|
|
|
|
|
my $smallest_bucket = $buckets[0]; |
714
|
|
|
|
|
|
|
|
715
|
|
|
|
|
|
|
return if ! defined $smallest_bucket || ! $smallest_bucket || ! @buckets; |
716
|
|
|
|
|
|
|
|
717
|
|
|
|
|
|
|
foreach my $date (@dates) { |
718
|
|
|
|
|
|
|
confess "Can't trim a non-DateTime value" unless $date->isa( 'DateTime' ); |
719
|
|
|
|
|
|
|
|
720
|
|
|
|
|
|
|
#foreach my $bucket (grep { $_->trimmable && ($_->order < $smallest_bucket->order) } values %DateTime::Event::Predict::Profile::BUCKETS) { |
721
|
|
|
|
|
|
|
foreach my $bucket (grep { $_->order < $smallest_bucket->order } values %DISTINCT_BUCKETS) { |
722
|
|
|
|
|
|
|
# Clone the date so we don't modify anything we shouldn't |
723
|
|
|
|
|
|
|
$date->clone->truncate( to => $smallest_bucket->accessor ); |
724
|
|
|
|
|
|
|
} |
725
|
|
|
|
|
|
|
} |
726
|
|
|
|
|
|
|
} |
727
|
|
|
|
|
|
|
|
728
|
|
|
|
|
|
|
# Useless syntactic sugar |
729
|
|
|
|
|
|
|
sub _trim_date { return &_trim_dates(@_); } |
730
|
|
|
|
|
|
|
|
731
|
|
|
|
|
|
|
1; # End of DateTime::Event::Predict |
732
|
|
|
|
|
|
|
|
733
|
|
|
|
|
|
|
__END__ |