From 6f269df9027a6289ed634c6160ebcacdb372f238 Mon Sep 17 00:00:00 2001
From: dakkar <dakkar@thenautilus.net>
Date: Sat, 28 Dec 2019 17:27:31 +0000
Subject: more types, counts seems to work better this way

---
 lib/MaildirIndexer/Index/ByAddresses.pm6 | 23 ++++++++++++-----------
 1 file changed, 12 insertions(+), 11 deletions(-)

diff --git a/lib/MaildirIndexer/Index/ByAddresses.pm6 b/lib/MaildirIndexer/Index/ByAddresses.pm6
index 4de80e8..e2a96c3 100644
--- a/lib/MaildirIndexer/Index/ByAddresses.pm6
+++ b/lib/MaildirIndexer/Index/ByAddresses.pm6
@@ -19,11 +19,11 @@ has $!total-count;
 method dump() {
 }
 
-submethod account-for(Str @addresses,$mailbox,Int $step) {
+submethod account-for(Str @addresses,Str $mailbox,Int $step) {
     $!total-count += $step;
     $!count-by-mailbox{$mailbox} += $step;
 
-    for @addresses -> $addr {
+    for @addresses -> Str $addr {
         $!known-addresses{$addr} += $step;
         $!count-by-address-and-mailbox{$addr => $mailbox} += $step;
     }
@@ -50,9 +50,9 @@ method del-path(IO:D $file, Str:D $mailbox --> Nil) {
     }
 }
 
-submethod p-address-given-mailbox($addr,$mailbox) {
-    my $a = 1 + ($!count-by-address-and-mailbox{$addr => $mailbox} // 0);
-    my $b = 2 + ($!count-by-mailbox{$mailbox} // 0);
+submethod p-address-given-mailbox(Str $addr, Str $mailbox) {
+    my $a = 1 + $!count-by-address-and-mailbox{$addr => $mailbox};
+    my $b = 2 + $!count-by-mailbox{$mailbox};
     return $a / $b;
 }
 
@@ -60,18 +60,19 @@ submethod predict-mailbox-given-addresses(@addresses) {
     my %prediction;
     my Bag $addr-bag .= new(|@addresses);
 
-    for $!count-by-mailbox.keys -> $mailbox {
-        my $p = 1;
+    for $!count-by-mailbox.keys -> Str $mailbox {
+        my $p = $!count-by-mailbox{$mailbox} / $!total-count;
 
-        for $!known-addresses.keys -> $addr {
+        for $!known-addresses.keys -> Str $addr {
+            my $addr-p = self.p-address-given-mailbox($addr,$mailbox);
             if ($addr-bag{$addr}) {
-                $p *= self.p-address-given-mailbox($addr,$mailbox);
+                $p *= $addr-p;
             }
             else {
-                $p *= (1 - self.p-address-given-mailbox($addr,$mailbox));
+                $p *= 1 - $addr-p;
             }
         }
-        $p *= $!count-by-mailbox{$mailbox} / $!total-count;
+
         %prediction{$mailbox} = $p;
     }
 
-- 
cgit v1.2.3