Creating Tableau Dashboards with Tableau desktop
The task was to create multiple dashboards from weather and tide data downloaded in Rstudio using multiple API's.
The weather and tide data was downloaded from the open-meteo.com API and the UK Admiralty marine data portal.
The data is is downloaded by 3 seperate R scripts
1. Current weather conditions using the open-meteo.com API.
2. A 7 day forecast using the open-meteo.com API.
3. A 6 day tidal forcast for a local site using the UK Admiralty marine data portal.
The 3 scripts can be scheduled to run on a windows machine using task scheduler to obtain regular updates as required. Unfortuently Tableau desktop does not automatically refresh the data and this has to be done manually by pressing F5 or clicking refresh on the Data tab.
The scripts created csv files which were used as data sources in Tableau. with relationships created using date time fields where required.
The first page was a simple dashboard to show the current weather. The idea was to have a easy to read layout with simple graphics. I toyed with a coloured background but was not happy with the results.
A calculated field was created to ensure that the rainfall displayed "0" when a null value appeared in the data and a new sheet was created to convert any of the 27 WMO weather codes to a description. Sunrise and sunset times were added at the bottom.
In hindsight the wind direction could have been better displayed as a description rather than a bearing. e.g. South westerly.
A second dashboard was created to display a general weather outlook. The aim was to provide a quick reference forecast for the upcomming days. This was done using the WMO Weather code values from the 7 day forecast. The created sheet was used again to tie the 27 different WMO codes to a descrition and display it alongside the code. This is dynamic and the chart will only display the values in the dataset.
A third dashboard showed a more detailed forecast. Again a filter was added to allow users to select the number of days to view. Forceast temperature and apparent temperature were displayed in the top chart with rainfail in the middle and humidity and pressure on the bottom. A trend line was added to show the temperature trend. Legends were displayed to show further details.
Forecasted rain was show in a bar chart and colours scheme added to highlight the amount of rain.
Pressure and humidity levels were displayed on a combined chart with different colours for each.
The next dashboard was solely about forecasted wind speed and direction. A graph showed the predicted wind speed and gusts while the bottom chart show the wind speed and direction. Arrows were used to show the direction of the wind and colours indicated the speed. Reference lines were also added on the Y axis to make it the direction trend a little easier to interpret.
The last dashboard was a simple tidal graph showing the high and low water times. Different colours were used to indicate the highs and lows and labels added to show the height. A continous line was created to link the high and low tide values.
As a bonus a Wind radial diagram was created to highlight the direction and strength of the wind over a time period. The arrows indicate the direction of the wind while the colour and size indicate the strength.
