MySQL three way join? - mysql

Question

I need to amend the following SQL so that it retrieves TTFF data for items that have category "beer" set in a table called planogram
select p.product_name, count(t.product_id), avg(t.TTFF), std(t.TTFF)
from product p left join
TTFFdata t
on p.product_id = t.product_id
where planogram_id = 1 group by p.product_name
However the products don't have category data, so I need to somehow get if from the categories in the planogram table. Something like this, but with the appropriate joining to make it work:
select p.product_name, count(t.product_id), avg(t.TTFF), std(t.TTFF)
from product p left join
TTFFdata t
on p.product_id = t.product_id
where planogram.category = "beer" group by p.product_name
Two things to consider:
I'm not very experiencd with SQL.
I think the database I've been given is badly structured (no category ids on the products?!), but I probably have to work with it as it is.
Any help much appreciated.
The full sql I'm working with is below:
-- phpMyAdmin SQL Dump
-- version 4.6.5.2
-- https://www.phpmyadmin.net/
--
-- Host: localhost
-- Generation Time: Mar 08, 2017 at 07:50 PM
-- Server version: 10.1.21-MariaDB
-- PHP Version: 5.6.30
SET SQL_MODE = "NO_AUTO_VALUE_ON_ZERO";
SET time_zone = "+00:00";
/*!40101 SET #OLD_CHARACTER_SET_CLIENT=##CHARACTER_SET_CLIENT */;
/*!40101 SET #OLD_CHARACTER_SET_RESULTS=##CHARACTER_SET_RESULTS */;
/*!40101 SET #OLD_COLLATION_CONNECTION=##COLLATION_CONNECTION */;
/*!40101 SET NAMES utf8mb4 */;
--
-- Database: `chart`
--
-- --------------------------------------------------------
--
-- Table structure for table `planogram`
--
CREATE TABLE `planogram` (
`planogram_id` int(11) NOT NULL,
`planogram_name` varchar(200) NOT NULL,
`project_name` varchar(200) NOT NULL,
`category` varchar(45) NOT NULL
) ENGINE=InnoDB DEFAULT CHARSET=latin1;
--
-- Dumping data for table `planogram`
--
INSERT INTO `planogram` (`planogram_id`, `planogram_name`, `project_name`, `category`) VALUES
(1, 'planogramA', '', 'beer'),
(2, 'planogramB', '', 'coffee'),
(3, 'planogramC', '', 'coffee'),
(4, 'planogramD', '', 'fruit'),
(5, 'planogramE', '', 'beer');
-- --------------------------------------------------------
--
-- Table structure for table `product`
--
CREATE TABLE `product` (
`product_id` int(11) NOT NULL,
`product_name` varchar(200) NOT NULL,
`detail` varchar(200) NOT NULL
) ENGINE=InnoDB DEFAULT CHARSET=latin1;
--
-- Dumping data for table `product`
--
INSERT INTO `product` (`product_id`, `product_name`, `detail`) VALUES
(1, 'beer1', ''),
(2, 'beer2', ''),
(3, 'coffee1', ''),
(4, 'coffee2', ''),
(5, 'coffee3', ''),
(6, 'coffee4', ''),
(7, 'coffee5', ''),
(8, 'coffee6', ''),
(9, 'beer3', ''),
(10, 'beer4', ''),
(13, 'fruit1', ''),
(14, 'fruit2', '');
-- --------------------------------------------------------
--
-- Table structure for table `TTFFdata`
--
CREATE TABLE `TTFFdata` (
`record_id` int(11) NOT NULL,
`participant_id` varchar(45) NOT NULL,
`TTFF` decimal(10,2) NOT NULL,
`product_id` int(11) NOT NULL,
`planogram_id` int(11) NOT NULL
) ENGINE=InnoDB DEFAULT CHARSET=latin1;
--
-- Dumping data for table `TTFFdata`
--
INSERT INTO `TTFFdata` (`record_id`, `participant_id`, `TTFF`, `product_id`, `planogram_id`) VALUES
(1, 'a1', '6.21', 1, 1),
(2, 'a2', '5.70', 1, 1),
(3, 'a3', '6.00', 1, 1),
(4, 'b1', '3.40', 2, 1),
(5, 'b2', '4.30', 2, 1),
(6, 'b3', '6.00', 2, 1),
(7, 'c1', '7.00', 3, 2),
(8, 'c2', '8.00', 3, 2),
(9, 'c3', '5.00', 3, 2),
(10, 'd1', '8.90', 4, 2),
(11, 'd2', '3.00', 4, 2),
(12, 'd3', '4.50', 4, 2),
(15, 'e1', '8.00', 5, 2),
(16, 'e2', '9.00', 5, 2),
(17, 'f1', '5.50', 6, 3),
(18, 'f2', '5.00', 6, 3),
(19, 'g1', '5.20', 7, 3),
(20, 'g2', '3.60', 7, 3),
(21, 'h1', '5.00', 8, 3),
(22, 'h2', '5.10', 8, 3),
(23, 'i1', '6.00', 13, 4),
(24, 'i2', '7.00', 13, 4),
(25, 'j1', '4.00', 14, 4),
(26, 'j2', '6.00', 14, 4),
(27, 'k1', '6.70', 9, 5),
(28, 'k2', '6.10', 9, 5),
(29, 'l1', '3.00', 10, 5),
(30, 'l2', '5.00', 10, 5);
--
-- Indexes for dumped tables
--
--
-- Indexes for table `planogram`
--
ALTER TABLE `planogram`
ADD PRIMARY KEY (`planogram_id`);
--
-- Indexes for table `product`
--
ALTER TABLE `product`
ADD PRIMARY KEY (`product_id`);
--
-- Indexes for table `TTFFdata`
--
ALTER TABLE `TTFFdata`
ADD PRIMARY KEY (`record_id`),
ADD KEY `product_id` (`product_id`),
ADD KEY `planogram_id` (`planogram_id`);
--
-- AUTO_INCREMENT for dumped tables
--
--
-- AUTO_INCREMENT for table `planogram`
--
ALTER TABLE `planogram`
MODIFY `planogram_id` int(11) NOT NULL AUTO_INCREMENT, AUTO_INCREMENT=6;
--
-- AUTO_INCREMENT for table `product`
--
ALTER TABLE `product`
MODIFY `product_id` int(11) NOT NULL AUTO_INCREMENT, AUTO_INCREMENT=17;
--
-- AUTO_INCREMENT for table `TTFFdata`
--
ALTER TABLE `TTFFdata`
MODIFY `record_id` int(11) NOT NULL AUTO_INCREMENT, AUTO_INCREMENT=31;
/*!40101 SET CHARACTER_SET_CLIENT=#OLD_CHARACTER_SET_CLIENT */;
/*!40101 SET CHARACTER_SET_RESULTS=#OLD_CHARACTER_SET_RESULTS */;
/*!40101 SET COLLATION_CONNECTION=#OLD_COLLATION_CONNECTION */;

Answer 1

You need to include all 3 tables in the join and make sure that you group by on all relevant fields to get the correct output. Based on the sample data, each product is associated with only one planogram, so my answer is based on the assumption that this is true for your entire data:
select p.product_name, count(t.product_id), avg(t.TTFF), std(t.TTFF)
from product p
left join TTFFdata t on p.product_id = t.product_id
left join planogram pl on t.planogram_id=pl.planogram_id
where planogram.category = "beer"
group by p.product_name