Automatic Calendar Generation

This document explains how to use the automatic calendar generation feature and understand its different options.

Introduction

The automatic calendar generation is a powerful tool that creates an optimized schedule based on your requirements, user preferences, and historical data. The system uses advanced algorithms to ensure fair distribution of shifts while meeting all specified constraints.

Accessing the Generator

1. Access the calendar as administrator
2. Click on the "Generate Calendar" button in the top-right
3. A modal window will appear with generation options

Basic Settings

Date Range

• Initial Date: Select the start date for the period you want to generate
• Final Date: Select the end date for the period
• Current Month: By default, the current month is selected
• Note: It's recommended to generate periods of 1-3 months for optimal results

Historical Data Options

Choose how historical data should be considered:

• Year to Date (YTD): Uses data from the beginning of the current year
• Last 12 Months: Uses data from the previous 12 months (recommended)
• No Historical Data: Generates the calendar without considering past assignments
• Custom Period: Allows you to specify a custom date range for historical data

Advanced Options

Click on "Advanced Options" to access additional settings that control how the generation algorithm works:

Optimization Weights

These settings determine the relative importance of different factors in the generation process:

1.Day type leveling (0-100%):

• Controls how much emphasis is placed on distributing different types of days evenly
• Higher values ensure more balanced distribution of workdays/weekends/bank holidays.
• Default: 100%

2.Demand type leveling (0-100%):

• Controls the emphasis on distributing shifts evenly across all demand types for each user
• Higher values ensure a more balanced distribution of demand types among users.
• Default: 100%

3.Distribution leveling (0-100%):

• Ensure that the demands are distributed as evenly as possible across the days. This means that workdays should not be clustered too closely together or spaced too far apart.
• Higher values ensure a more equal distribution.
• Default: 100%.

Note: By default, all options have same weight. If you want turn on only one option, you can set the other options to 0%.

Generation Process

Process Itinerary:

• The system validates all input data, checks for any conflicts in user preferences or demands, and verifies the availability of historical data.
• It forecasts whether it can resolve the problem within a reasonable timeframe.
• It uses the optimization algorithm to solve the problem.
• It calculates statistics for the solution.

Three Possible Outcomes:

Solution (green banner):

• The algorithm finds a solution that meets all constraints and optimizes the specified criteria.
• Click on "Details" to view the solution statistics.
• You can apply the solution to the calendar by clicking "Apply."

Problem Detected (red banner):

• There are issues with the input data. Errors will be displayed, and the process will be halted.
• We recommend correcting the errors and regenerating the calendar.

Suboptimal Solution (yellow banner):

• The algorithm did not accurately assess the difficulty of the problem and allocated fewer resources than necessary. Nevertheless, a solution has been found, but it may not be the best one.
• Click on "Details" to view the solution statistics.
• You can preview the solution by clicking "Preview."
• You can keep the solution by clicking "Apply."
• You can have the algorithm attempt to find a better solution by clicking "Improve."

Backup and Data Cleanup:

• If you are not satisfied with the applied solution, you can restore the last backup by clicking "Revert to Previous Version" in the buttons at the top right.
• If you simply want to delete what the model has generated without reverting to a previous state, you can click "Clear Data" in the dropdown menu at the top right.

Best Practices

1.Historical Data:

• Using historical data helps create more balanced schedules
• The "Last 12 Months" option is recommended for most cases
• Consider using "No Historical Data" only when starting fresh

2.Optimization Weights:

• Start with default values (100%)
• Adjust weights if specific aspects need more attention
• Make small adjustments and test results

3.Generation Period:

• Generate 1-3 months at a time for best results
• Longer periods may require more processing time
• Consider user preferences and holiday periods

Troubleshooting

Common issues

• Missing Demand For Fixed Events: There is no demand for the fixed events assigned to the users.
• Total Demand Is Zero: Total demand must be greater than zero.
• Number Of Users Is Zero: The number of users must be greater than zero.
• Demand Is Greater Than Users Capacity: Demand exceeds the capacity of the users.
• No Polivalency for Fixed Events: Any users with fixed event has no polivalency for the demand.
• There Is No User To Cover Demand: There is no user available to cover the demand.
• Demand Greater Than Users Capacity Specific: The demand for demand name exceeds the capacity of the users who can cover it. Please reduce this type of demand or increase the number of users who can cover it.
• Demand Greater Than Users Capacity General: There are not enough users to cover all tasks.
• More Demands Than Users: There are more demands than different users on one date.

Limits

Nowadays, we have some limits to avoid too very complex problems that may take too much time to solve:

• Max Days per Problem: There is a limit of days to generate in one go. You can split the problem in smaller periods.
• Very High Complexity: The problem is too complex to solve. Please try simplifying the problem (maybe you can reduce the number of users or demands).

Note: If you have any issues, you can contact us by clicking on the following link Contact us.