'.
+ &mt('Press "Generate Statistics" when you are ready.').
+ '
'.
+ &mt('It may take some time to update the student data '.
+ 'for the first analysis. Future analysis this session '.
+ ' will not have this delay.').
+ '
');
return;
+ } elsif ($ENV{'form.statsfirstcall'} eq 'yes' ||
+ exists($ENV{'form.UpdateCache'}) ||
+ exists($ENV{'form.ClearCache'}) ) {
+ $r->print('');
+ &Apache::lonstatistics::Gather_Student_Data($r);
+ } else {
+ $r->print('');
}
+ $r->rflush();
#
- &Apache::lonstatistics::Gather_Student_Data($r);
- #
+ # This probably does not need to be done each time we are called, but
+ # it does not slow things down noticably.
+ &Apache::loncoursedata::populate_weight_table();
#
- if ($output_mode eq 'html') {
- $r->print("
\n");
- my ($starttime,$endtime) = &Apache::lonstathelpers::get_time_limits();
- if (defined($starttime) || defined($endtime)) {
- # Inform the user what the time limits on the data are.
- $r->print('
'.&mt('Statistics on submissions from [_1] to [_2]',
- &Apache::lonlocal::locallocaltime($starttime),
- &Apache::lonlocal::locallocaltime($endtime)).
- '
');
- }
- $r->print("
".&mt('Compiled on [_1]',
- &Apache::lonlocal::locallocaltime(time))."
\n");
+ return;
+}
-Presents the statistics data as an html table organized by the order
-the assessments appear in the course.
+sub html_preamble {
+ my $Str='';
+ $Str .= "
\n";
+ my ($starttime,$endtime) = &Apache::lonstathelpers::get_time_limits();
+ if (defined($starttime) || defined($endtime)) {
+ # Inform the user what the time limits on the data are.
+ $Str .= '
'.&mt('Statistics on submissions from [_1] to [_2]',
+ &Apache::lonlocal::locallocaltime($starttime),
+ &Apache::lonlocal::locallocaltime($endtime)
+ ).'
';
+ }
+ $Str .= "
".&mt('Compiled on [_1]',
+ &Apache::lonlocal::locallocaltime(time))."
";
+ return $Str;
+}
-=cut
###############################################
###############################################
-sub output_html_grouped_by_sequence {
- my ($r) = @_;
- my $problem_num = 0;
- #$r->print(&ProblemStatisticsLegend());
- foreach my $sequence (&Apache::lonstatistics::Sequences_with_Assess()) {
- next if ($sequence->{'num_assess'}<1);
- $r->print("
".$sequence->{'title'}."
");
- $r->print('
'."\n");
- $r->print('
'."\n");
- $r->print('
');
- my $Str = &statistics_table_header('no container no plots');
- $r->print('
'.$Str."
\n");
- foreach my $resource (@{$sequence->{'contents'}}) {
- next if ($resource->{'type'} ne 'assessment');
- foreach my $part (@{$resource->{'parts'}}) {
- $problem_num++;
- my $data = &get_statistics($sequence,$resource,$part,
- $problem_num);
- my $option = '';
- $r->print('
';
+ foreach my $field (@SeqFields) {
+# next if ($field->{'selected'} ne 'yes');
+ $Str .= '
{'title'}.'
';
+ }
+ $Str .= '
';
+ return $Str;
+}
-=item &output_html_ungrouped()
-Presents the statistics data in a single html table which can be sorted by
-different columns.
+sub sequence_html_output {
+ my ($seq) = @_;
+ my $data = $SeqStat{$seq->{'symb'}};
+ my $row = '
';
+ foreach my $field (@SeqFields) {
+ next if ($field->{'selected'} ne 'yes');
+ $row .= '
';
+ ${$cols_output}+= 1;
}
- return $row;
+ ${$cols_output} =0;
+ ${$rows_output}+=1;
+ ##
+ ## Then the short titles
+ foreach my $field (@{$Fields}) {
+ next if ($field->{'selected'} ne 'yes');
+ next if ($field->{'name'} eq 'problem_num');
+ # Use english for excel as I am not sure how well excel handles
+ # other character sets....
+ $excel_sheet->write($$rows_output,$$cols_output,
+ $field->{'title'},
+ $format->{'bold'});
+ $$cols_output+=1;
+ }
+ ${$cols_output} =0;
+ ${$rows_output}+=1;
+ return;
}
-sub statistics_table_header {
- my ($options) = @_;
- my $header_row;
- foreach my $field (@Fields) {
- next if ($options =~ /no $field->{'name'}/);
- $header_row .= '
';
- if ($options =~ /sortable/ &&
- exists($field->{'sortable'}) && $field->{'sortable'} eq 'yes') {
- $header_row .= '{'name'}."'".
- ';document.Statistics.submit();">';
- }
- $header_row .= &mt($field->{'title'});
- if ($options =~ /sortable/) {
- $header_row.= '';
- }
- if ($options !~ /no plots/ &&
- exists($field->{'graphable'}) &&
- $field->{'graphable'} eq 'yes') {
- $header_row.=' (';
- $header_row .= '';
- $header_row .= &mt('plot').')';
+##################################################
+##################################################
+##
+## Statistics Gathering and Manipulation Routines
+##
+##################################################
+##################################################
+sub compute_statistics_on_sequence {
+ my ($seq) = @_;
+ my @Data;
+ foreach my $res (@{$seq->{'contents'}}) {
+ next if ($res->{'type'} ne 'assessment');
+ foreach my $part (@{$res->{'parts'}}) {
+ #
+ # This is where all the work happens
+ my $data = &get_statistics($seq,$res,$part,scalar(@StatsArray)+1);
+ push (@Data,$data);
+ push (@StatsArray,$data);
}
- $header_row .= '
';
}
- return $header_row;
+ return @Data;
}
-###############################################
-###############################################
-
-=pod
-
-=item &plot_statistics()
-
-=cut
+sub compute_all_statistics {
+ my ($r) = @_;
+ if (@StatsArray > 0) {
+ # Assume we have already computed the statistics
+ return;
+ }
+ my $c = $r->connection;
+ foreach my $seq (&Apache::lonstatistics::Sequences_with_Assess()) {
+ last if ($c->aborted);
+ next if ($seq->{'num_assess'} < 1);
+ &compute_sequence_statistics($seq);
+ &compute_statistics_on_sequence($seq);
+ }
+}
-###############################################
-###############################################
-sub plot_statistics {
- my ($r,$datafield) = @_;
- my @Data;
+sub sort_data {
+ my ($sortkey) = @_;
+ return if (! @StatsArray);
#
- #
- my $sortfield = undef;
- my $title = undef;
+ # Sort the data
+ my $sortby = undef;
foreach my $field (@Fields) {
- if ($datafield eq $field->{'name'} &&
- exists($field->{'graphable'}) && $field->{'graphable'} eq 'yes') {
- $sortfield = $field->{'name'};
- $title = $field->{'long_title'};
- }
- }
- return if (! defined($sortfield) || $sortfield eq '');
- #
- my $Max = 0;
- my $problem_num = 0;
- foreach my $sequence (&Apache::lonstatistics::Sequences_with_Assess()) {
- next if ($sequence->{'num_assess'}<1);
- foreach my $resource (@{$sequence->{'contents'}}) {
- next if ($resource->{'type'} ne 'assessment');
- foreach my $part (@{$resource->{'parts'}}) {
- my $problem_number++;
- my $data = &get_statistics($sequence,$resource,$part,
- $problem_num);
- my $value = $data->{$sortfield};
- $Max = $value if ($Max < $value);
- push (@Data,$value);
- }
+ if ($sortkey eq $field->{'name'}) {
+ $sortby = $field->{'name'};
}
}
- #
- # Print out plot request
- my $yaxis = '';
- if ($sortfield eq 'per_wrong') {
- $yaxis = 'Percent';
- }
- #
- # Determine appropriate value for $Max
- if ($sortfield eq 'deg_of_diff') {
- if ($Max > 0.5) {
- $Max = 1;
- } elsif ($Max > 0.2) {
- $Max = 0.5;
- } elsif ($Max > 0.1) {
- $Max = 0.2;
- }
- } elsif ($sortfield eq 'per_wrong') {
- if ($Max > 50) {
- $Max = 100;
- } elsif ($Max > 25) {
- $Max = 50;
- } elsif ($Max > 20) {
- $Max = 25;
- } elsif ($Max > 10) {
- $Max = 20;
- } elsif ($Max > 5) {
- $Max = 10;
+ if (! defined($sortby) || $sortby eq '' || $sortby eq 'problem_num') {
+ $sortby = 'container';
+ }
+ if ($sortby ne 'container') {
+ # $sortby is already defined, so we can charge ahead
+ if ($sortby =~ /^(title|part)$/i) {
+ # Alpha comparison
+ @StatsArray = sort {
+ lc($a->{$sortby}) cmp lc($b->{$sortby}) ||
+ lc($a->{'title'}) cmp lc($b->{'title'}) ||
+ lc($a->{'part'}) cmp lc($b->{'part'});
+ } @StatsArray;
} else {
- $Max = 5;
+ # Numerical comparison
+ @StatsArray = sort {
+ my $retvalue = 0;
+ if ($b->{$sortby} eq 'nan') {
+ if ($a->{$sortby} ne 'nan') {
+ $retvalue = -1;
+ } else {
+ $retvalue = 0;
+ }
+ }
+ if ($a->{$sortby} eq 'nan') {
+ if ($b->{$sortby} ne 'nan') {
+ $retvalue = 1;
+ }
+ }
+ if ($retvalue eq '0') {
+ $retvalue = $b->{$sortby} <=> $a->{$sortby} ||
+ lc($a->{'title'}) <=> lc($b->{'title'}) ||
+ lc($a->{'part'}) <=> lc($b->{'part'});
+ }
+ $retvalue;
+ } @StatsArray;
}
}
-
- $r->print("
\n");
#
- # Print out the data
- $ENV{'form.sortby'} = 'Contents';
-# &output_html_ungrouped($r);
+ # Renumber the data set
+ my $count;
+ foreach my $data (@StatsArray) {
+ $data->{'problem_num'} = ++$count;
+ }
return;
}
@@ -831,7 +1463,8 @@ sub plot_statistics {
=item &get_statistics()
Wrapper routine from the call to loncoursedata::get_problem_statistics.
-Calls lonstathelpers::get_time_limits() to limit the data set by time.
+Calls lonstathelpers::get_time_limits() to limit the data set by time
+and &compute_discrimination_factor
Inputs: $sequence, $resource, $part, $problem_num
@@ -853,6 +1486,7 @@ sub get_statistics {
(\@Apache::lonstatistics::SelectedSections,
$Apache::lonstatistics::enrollment_status,
$symb,$part,$courseid,$starttime,$endtime);
+ $data->{'symb'} = $symb;
$data->{'part'} = $part;
$data->{'problem_num'} = $problem_num;
$data->{'container'} = $sequence->{'title'};
@@ -860,11 +1494,34 @@ sub get_statistics {
$data->{'title.link'} = $resource->{'src'}.'?symb='.
&Apache::lonnet::escape($resource->{'symb'});
#
- $data->{'deg_of_disc'} = &compute_discrimination_factor($resource,$part,$sequence);
+ if ($SelectedFields{'deg_of_disc'}) {
+ $data->{'deg_of_disc'} =
+ &compute_discrimination_factor($resource,$part,$sequence);
+ }
+ #
+ # Store in metadata if computations were done for all students
+ if ($data->{'num_students'} > 1) {
+ my @Sections = @Apache::lonstatistics::SelectedSections;
+ my $sections = '"'.join(' ',@Sections).'"';
+ $sections =~ s/&+/_/g; # Ensure no special characters
+ $data->{'sections'}=$sections;
+ $data->{'course'} = $ENV{'request.course.id'};
+ my $urlres=(&Apache::lonnet::decode_symb($resource->{'symb'}))[2];
+ $data->{'urlres'}=$urlres;
+ my %storestats =
+ &LONCAPA::lonmetadata::dynamic_metadata_storage($data);
+ my ($dom,$user) = $urlres=~/^(\w+)\/(\w+)/;
+ &Apache::lonnet::put('nohist_resevaldata',\%storestats,$dom,$user);
+ }
+ #
+ # Get the due date for research purposes (commented out most of the time)
+# $data->{'duedate'} =
+# &Apache::lonnet::EXT('resource.'.$part.'.duedate',$symb);
+# $data->{'opendate'} =
+# &Apache::lonnet::EXT('resource.'.$part.'.opendate',$symb);
return $data;
}
-
###############################################
###############################################
@@ -882,7 +1539,6 @@ Returns: integer between -1 and 1
###############################################
sub compute_discrimination_factor {
my ($resource,$part,$sequence) = @_;
- &Apache::lonnet::logthis($sequence->{'title'}.' '.$resource->{'title'});
my @Resources;
foreach my $res (@{$sequence->{'contents'}}) {
next if ($res->{'symb'} eq $resource->{'symb'});
@@ -890,11 +1546,13 @@ sub compute_discrimination_factor {
}
#
# rank
+ my ($starttime,$endtime) = &Apache::lonstathelpers::get_time_limits();
my $ranking =
&Apache::loncoursedata::rank_students_by_scores_on_resources
(\@Resources,
\@Apache::lonstatistics::SelectedSections,
- $Apache::lonstatistics::enrollment_status,undef);
+ $Apache::lonstatistics::enrollment_status,undef,
+ $starttime,$endtime);
#
# compute their percent scores on the problems in the sequence,
my $number_to_grab = int(scalar(@{$ranking})/4);
@@ -906,9 +1564,11 @@ sub compute_discrimination_factor {
$_->[&Apache::loncoursedata::RNK_student()];
} @{$ranking}[($num_students-$number_to_grab)..($num_students-1)];
my ($bottom_sum,$bottom_max) =
- &Apache::loncoursedata::get_sum_of_scores($resource,$part,\@BottomSet);
+ &Apache::loncoursedata::get_sum_of_scores($resource,$part,\@BottomSet,
+ undef,$starttime,$endtime);
my ($top_sum,$top_max) =
- &Apache::loncoursedata::get_sum_of_scores($resource,$part,\@TopSet);
+ &Apache::loncoursedata::get_sum_of_scores($resource,$part,\@TopSet,
+ undef,$starttime,$endtime);
my $deg_of_disc;
if ($top_max == 0 || $bottom_max==0) {
$deg_of_disc = 'nan';
@@ -922,73 +1582,132 @@ sub compute_discrimination_factor {
###############################################
###############################################
+##
+## Compute KR-21
+##
+## To compute KR-21, you need the following information:
+##
+## K=the number of items in your test
+## M=the mean score on the test
+## s=the standard deviation of the scores on your test
+##
+## then:
+##
+## KR-21 rk= [K/(K-1)] * [1- (M*(K-M))/(K*s^2))]
+##
+###############################################
+###############################################
+sub compute_sequence_statistics {
+ my ($seq) = @_;
+ my $symb = $seq->{'symb'};
+ my @Resources;
+ foreach my $res (@{$seq->{'contents'}}) {
+ next if ($res->{'type'} ne 'assessment');
+ push (@Resources,$res->{'symb'});
+ }
+ my ($starttime,$endtime) = &Apache::lonstathelpers::get_time_limits();
+ #
+ # First compute statistics based on student scores
+ my ($smin,$smax,$sMean,$sSTD,$scount,$sMAX) =
+ &Apache::loncoursedata::score_stats
+ (\@Apache::lonstatistics::SelectedSections,
+ $Apache::lonstatistics::enrollment_status,
+ \@Resources,$starttime,$endtime,undef);
+ $SeqStat{$symb}->{'title'} = $seq->{'title'};
+ $SeqStat{$symb}->{'scoremax'} = $smax;
+ $SeqStat{$symb}->{'scoremin'} = $smin;
+ $SeqStat{$symb}->{'scoremean'} = $sMean;
+ $SeqStat{$symb}->{'scorestd'} = $sSTD;
+ $SeqStat{$symb}->{'scorecount'} = $scount;
+ $SeqStat{$symb}->{'max_possible'} = $sMAX;
+ #
+ # Compute statistics based on the number of correct problems
+ # 'correct' is taken to mean
+ my ($cmin,$cmax,$cMean,$cSTD,$ccount)=
+ &Apache::loncoursedata::count_stats
+ (\@Apache::lonstatistics::SelectedSections,
+ $Apache::lonstatistics::enrollment_status,
+ \@Resources,$starttime,$endtime,undef);
+ my $K = $seq->{'num_assess_parts'};
+ my $kr_21;
+ if ($K > 1 && $cSTD > 0) {
+ $kr_21 = ($K/($K-1)) * (1 - $cMean*($K-$cMean)/($K*$cSTD**2));
+ } else {
+ $kr_21 = 'nan';
+ }
+ $SeqStat{$symb}->{'countmax'} = $cmax;
+ $SeqStat{$symb}->{'countmin'} = $cmin;
+ $SeqStat{$symb}->{'countstd'} = $cSTD;
+ $SeqStat{$symb}->{'countmean'} = $cMean;
+ $SeqStat{$symb}->{'count'} = $ccount;
+ $SeqStat{$symb}->{'items'} = $K;
+ $SeqStat{$symb}->{'KR-21'}=$kr_21;
+ return;
+}
+
+
=pod
-=item &ProblemStatisticsLegend()
+=item ProblemStatisticsLegend
-HELP This needs to be localized, or at least generated automatically.
+=over 4
-=cut
+=item #Stdnts
+Total number of students attempted the problem.
-###############################################
-###############################################
-sub ProblemStatisticsLegend {
- my $Ptr = '';
- $Ptr = '
';
- $Ptr .= '
';
- $Ptr .= '#Stdnts
';
- $Ptr .= '
Total number of students attempted the problem.';
- $Ptr .= '
';
- $Ptr .= 'Tries
';
- $Ptr .= '
Total number of tries for solving the problem.';
- $Ptr .= '
';
- $Ptr .= 'Max Tries
';
- $Ptr .= '
Largest number of tries for solving the problem by a student.';
- $Ptr .= '
';
- $Ptr .= 'Mean
';
- $Ptr .= '
Average number of tries. [ Tries / #Stdnts ]';
- $Ptr .= '
';
- $Ptr .= '#YES
';
- $Ptr .= '
Number of students solved the problem correctly.';
- $Ptr .= '
';
- $Ptr .= '#yes
';
- $Ptr .= '
Number of students solved the problem by override.';
- $Ptr .= '
';
- $Ptr .= '%Wrong
';
- $Ptr .= '
Percentage of students who tried to solve the problem ';
- $Ptr .= 'but is still incorrect. [ 100*((#Stdnts-(#YES+#yes))/#Stdnts) ]';
- $Ptr .= '
';
- $Ptr .= 'DoDiff
';
- $Ptr .= '
Degree of Difficulty of the problem. ';
- $Ptr .= '[ 1 - ((#YES+#yes) / Tries) ]';
- $Ptr .= '
';
- $Ptr .= 'S.D.
';
- $Ptr .= '
Standard Deviation of the tries. ';
- $Ptr .= '[ sqrt(sum((Xi - Mean)^2)) / (#Stdnts-1) ';
- $Ptr .= 'where Xi denotes every student\'s tries ]';
- $Ptr .= '
';
- $Ptr .= 'Skew.
';
- $Ptr .= '
Skewness of the students tries.';
- $Ptr .= '[(sqrt( sum((Xi - Mean)^3) / #Stdnts)) / (S.D.^3)]';
- $Ptr .= '
';
- $Ptr .= 'Dis.F.
';
- $Ptr .= '
Discrimination Factor: A Standard for evaluating the ';
- $Ptr .= 'problem according to a Criterion ';
- $Ptr .= '[Criterion to group students into %27 Upper Students - ';
- $Ptr .= 'and %27 Lower Students] ';
- $Ptr .= '1st Criterion for Sorting the Students: ';
- $Ptr .= 'Sum of Partial Credit Awarded / Total Number of Tries ';
- $Ptr .= '2nd Criterion for Sorting the Students: ';
- $Ptr .= 'Total number of Correct Answers / Total Number of Tries';
- $Ptr .= '
';
- $Ptr .= '
Disc.
';
- $Ptr .= '
Number of Students had at least one discussion.';
- $Ptr .= '
';
- return $Ptr;
-}
+=item Tries
+Total number of tries for solving the problem.
+
+=item Max Tries
+Largest number of tries for solving the problem by a student.
+
+=item Mean
+Average number of tries. [ Tries / #Stdnts ]
+
+=item #YES
+Number of students solved the problem correctly.
+
+=item #yes
+Number of students solved the problem by override.
+
+=item %Wrong
+Percentage of students who tried to solve the problem
+but is still incorrect. [ 100*((#Stdnts-(#YES+#yes))/#Stdnts) ]
+
+=item DoDiff
+Degree of Difficulty of the problem.
+[ 1 - ((#YES+#yes) / Tries) ]
+
+=item S.D.
+Standard Deviation of the tries.
+[ sqrt(sum((Xi - Mean)^2)) / (#Stdnts-1)
+where Xi denotes every student\'s tries ]
+
+=item Skew.
+Skewness of the students tries.
+[(sqrt( sum((Xi - Mean)^3) / #Stdnts)) / (S.D.^3)]
+
+=item Dis.F.
+Discrimination Factor: A Standard for evaluating the
+problem according to a Criterion
+
+=item [Criterion to group students into %27 Upper Students -
+and %27 Lower Students]
+1st Criterion for Sorting the Students:
+Sum of Partial Credit Awarded / Total Number of Tries
+2nd Criterion for Sorting the Students:
+Total number of Correct Answers / Total Number of Tries
+
+=item Disc.
+Number of Students had at least one discussion.
+
+=back
+
+=cut
-#---- END Problem Statistics Web Page ----------------------------------------
+############################################################
+############################################################
1;
__END__