Add all values in array for each ID in Perl - arrays

Question

I have this table:
NAME |12/31/2016|VALUE
AAA |1/31/2017 |10
AAA |2/1/2017 |20
AAA |2/2/2017 |30
AAA |2/3/2017 |40
AAA |2/4/2017 |50
NAME |2/9/2017 |VALUE
BBB |2/10/2017 |20
BBB |2/11/2017 |30
BBB |2/12/2017 |40
BBB |2/13/2017 |50
BBB |2/14/2017 |60
and this would be my desired output:
NAME |DATE |VALUE
AAA |12/31/2016 |150
AAA |1/31/2017 |140
AAA |2/1/2017 |120
NAME |DATE |VALUE
BBB |2/9/2017 |200
BBB |2/10/2017 |180
BBB |2/11/2017 |150
What I want to do is, for each of the valid symbols, (AAA, BBB) I want to have three rows.
For the first row of each column, I want all the values added,
For example, row 1 value for AAA:
10+20+30+40+50 = 150
then for row 2 I want to just add from the second value to the last.
For example row 2 value for AAA
20+30+40+50 = 140
and so on same goes for BBB.
I want to shift the dates down so that 12/31/2016 would match AAA, then get the first three dates for each row.
I currently have this code. but this doesn't do much. it just gives me a bunch of numbers.
use strict;
use warnings;
use Scalar::Util qw(looks_like_number);
use Data::Dumper;
sub uniq {
my %seen;
grep !$seen{$_}++, #_;
}
my %cashflow;
my %fields = (
ID => 0,
DATES => 1,
VALUE => 2,
);
my #total;
my #IDs;
my #uniqueIDs;
my #dates;
my #add;
my $i = 0;
my #values;
my $counter = 3;
open( FILE, "try.CSV" );
while ( my $line = <FILE> ) {
chomp( $line );
my #lineVals = split( /\|/, $line );
if ( $lineVals[ $fields{ID} ] !~ /^SYMBOL$/i ) {
push #IDs, $lineVals[ $fields{ID} ];
}
#uniqueIDs = uniq( #IDs );
#push all CASH FLOW AMOUNTS to #cashflow
if ( looks_like_number( $lineVals[ $fields{VALUE} ] ) ) {
$lineVals[ $fields{VALUE} ] =~ s/\r//;
push #total, $lineVals[ $fields{VALUE} ];
}
if ( $lineVals[ $fields{DATES} ] =~ /(\d{1,2})\/(\d{1,2})\/(\d{4})/ ) {
$lineVals[ $fields{DATES} ] = sprintf( '%04d%02d%02d', $3, $2, $1 );
}
$cashflow{ uc $lineVals[ $fields{ID} ] }{DATES} = $lineVals[ $fields{DATES} ];
$cashflow{ uc $lineVals[ $fields{ID} ] }{VALUE} = $lineVals[ $fields{VALUE} ];
foreach my $ID ( #uniqueIDs ) {
foreach my $symb ( keys %cashflow ) {
if ( $ID = $symb ) {
if ( looks_like_number( $lineVals[ $fields{VALUE} ] ) ) {
$lineVals[ $fields{VALUE} ] =~ s/\r//;
push #total, $lineVals[ $fields{VALUE} ];
my $i = 0;
my $grand = 0;
foreach my $val ( #total ) {
while ( $i < $counter ) {
$grand += $val;
print "$grand \n";
$i++;
}
shift #total;
}
}
}
}
}
}
close FILE;
I'm really stuck with this. I don't know what to do with the problem.

Answer 1

A possible solution:
#!perl
use strict;
use warnings;
sub trim {
my ($str) = #_;
s!\A\s+!!, s!\s+\z!! for $str;
$str
}
my $file = 'try.CSV';
open my $fh, '<', $file or die "$0: $file: $!\n";
my ($group_name, #dates, #values);
my $sum = 0;
my $print_group = sub {
return if !defined $group_name;
my $format = " %-6s|%-11s|%s\n";
printf $format, 'NAME', 'DATE', 'VALUE';
for my $date (#dates) {
printf $format, $group_name, $date, $sum;
$sum -= shift #values if #values;
}
};
while (my $line = readline $fh) {
my ($name, $date, $value) = map trim($_), split /\|/, $line;
if ($name eq 'NAME') {
$print_group->();
$group_name = undef;
#dates = $date;
#values = ();
$sum = 0;
next;
}
$group_name ||= $name;
push #dates, $date if #dates < 3;
push #values, $value if #values < 2;
$sum += $value;
}
$print_group->();
Let's go over it.
sub trim {
my ($str) = #_;
s!\A\s+!!, s!\s+\z!! for $str;
$str
}
A helper function for removing leading/trailing whitespace from a string. We're using ! as the s delimiter here because / breaks SO's syntax highlighting. Shrug.
my $file = 'try.CSV';
open my $fh, '<', $file or die "$0: $file: $!\n";
Open our input file. Note: We use a lexical variable ($fh) instead of a bareword filehandle, and we use 3-argument open. This is strongly recommended. We also check open's return value and produce a nice error message in case of failure, including both the name of the file that couldn't be opened ($file) and the reason for failing ($!).
my ($group_name, #dates, #values);
my $sum = 0;
We set up some state variables that we want to preserve across loop iterations. $group_name is the name of the group we're currently processing, #dates is the saved dates we've seen so far, #values is the saved values we've seen so far. $sum is a running sum of all the values in the current group, and it starts at 0.
my $print_group = sub {
return if !defined $group_name;
my $format = " %-6s|%-11s|%s\n";
printf $format, 'NAME', 'DATE', 'VALUE';
for my $date (#dates) {
printf $format, $group_name, $date, $sum;
$sum -= shift #values if #values;
}
};
A helper function for printing the output for a single group. If $group_name isn't set, we haven't processed any input for the current group yet, so we do nothing and return. Otherwise we print a NAME | DATE | VALUE header, followed by a row of data for each element in #dates. For each $date we output the current group name (e.g. AAA), $date, and the sum of values (all nicely formatted using printf). Initially $sum is the sum of all group values, but after the first iteration we start subtracting the values from #values: If the list of values in the input was x1, x2, x3, x4, ..., then $sum is initially x1 + x2 + x3 + x4 + ..., and that's what's printed in the first line of output. After that we subtract x1, so the next line gets x1 + x2 + x3 + x4 + ... - x1, which is x2 + x3 + x4 + .... After that we subtract x2, so the third row of data gets x3 + x4 + ....
while (my $line = readline $fh) {
my ($name, $date, $value) = map trim($_), split /\|/, $line;
Our main loop. We read a line of input, split it on |, and trim each field.
if ($name eq 'NAME') {
$print_group->();
$group_name = undef;
#dates = $date;
#values = ();
$sum = 0;
next;
}
If $name is 'NAME', this is the start of a new group. Print the output for the current group if any ($print_group->() does nothing if there is no current group), then reset our state variables back to initial values, except for #dates, which is filled with the $date value from the header row. Then start the next iteration of the loop because we're done with this line.
$group_name ||= $name;
push #dates, $date if #dates < 3;
push #values, $value if #values < 2;
$sum += $value;
If we get here, this line is not the start of a new group. We set $group_name if it hasn't been set yet. We add $date to our list of saved dates (but we only need 3 dates, so do nothing if we already have 3). We add $value to our list of saved values (but we only need 2 of them). Finally we add $value to our total $sum within the group.
}
$print_group->();
At the end of the loop we've also just finished processing a group, so we need to call $print_group here as well.

Answer 2

This will do as you ask. It reads the whole data file into an array of arrays and manipulates that array before printing it. The blocks are processed backwards from the end so that the other blocks remain in place when the trailing lines are deleted
This program expects the path to the input file as a parameter on the command line and writes the result to STDOUT
use strict;
use warnings 'all';
my #data = map [ /[^|\s]+/g ], <>;
# Make a list of the indices of all the header rows
my #headers = grep { $data[$_][0] eq 'NAME' } 0 .. $#data;
# Make a list of the indices of the first
# and last lines of all the data blocks
my #blocks = map {
[
$headers[$_] + 1,
$_ == $#headers ? $#data : $headers[$_+1] - 1
]
} 0 .. $#headers;
# Shift the second column down
# Replace the col2 header with 'DATE'
#
$data[$_][1] = $data[$_-1][1] for reverse 1 .. $#data;
$data[$_][1] = 'DATE' for #headers;
# Edit each block of data
#
for my $block ( reverse #blocks ) {
my ( $beg, $end ) = #$block;
# Calculate the block total
my $total = 0;
for ( $beg ... $end ) {
$total += $data[$_][2];
}
# Calculate the first three data values
for my $i ( $beg .. $beg + 2 ) {
my $next = $total - $data[$i][2];
$data[$i][2] = $total;
$total = $next;
}
# Remove everything except those three lines
splice #data, $beg+3, $end-$beg-2;
}
print join('|', #$_), "\n" for #data;
output
NAME|DATE|VALUE
AAA|12/31/2016|150
AAA|1/31/2017|140
AAA|2/1/2017|120
NAME|DATE|VALUE
BBB|2/9/2017|200
BBB|2/10/2017|180
BBB|2/11/2017|150