So, it's my first time trying to make my own data visualization, and what I want to do is something like a heat map or highlighted table like the original Kohonen (but using squares rather than hexagons) research about countries, but I don't know how to apply it to the map I've got from training the network.
I've read a few links about making a highlighted table and heat map on Tableau, but what I get is always just a table mainly because I'm not using a measure (i just want every country of one color), so, my question is: is it possible to use Tableau for this situation and if so, how could I do it? Thanks!
Original SOM visualization from Dr. Kohonen
Table from Tableau
Yes it is possible to make visualizations like you describe in Tableau - but the question you posed is a bit vaguely specified to give you much more detailed guidance than first take some Tableau training.
For instance, how do you want to determine which row and column to place a country? You may need to write a table calc for row and col positions, which means learning about table calcs. Or you can assign them in a data column.
For square country marks, you'll want a discrete field on the rows shelf, a discrete field on the columns shelf, possible a dimension like Country name on the details shelf. Choose a square mark type, and (important) choose a square cell size from the format menu. Then adjust the size of your squares from the size button and the format menu, and colors and borders from the color button and format pane.
Lots to play with.
Related
I have created a dual axis chart where I'm graphing the arrival time and the departure time in the same graph, and showing the sum of the number of records as the rows.
By default Tableau adds the "Measure Names" into each of my columns Marks and I get an automated Legend showing the "Number of Records" in one colour.
However, this is not useful at all...
I want each time variable (eg. arrival time) to show up in a different colour. The only way I have been able to do this is by removing the "Measure Names" pill from the Marks card and then setting each axis to have its own colour. But as soon as I do this, the Legend disappears and now I finally get a nice graph with a line in each colour, but no legend to show what line is assigned to what colour.
Also the Legend option gets disabled when I use this layout.
How can I add the Legend to this?
Please note that the solution in this post doesn't help to answer this question.
You should be able to create 2 calculated fields that basically mimic Number of Records like
SUM([Number of Records])
Once you have 2 calculated fields, you can name them each what you desire. Then you should be able to use measure names and those separate calculated fields on each card of your dual axis.
Many, but not necessarily all, views people want to make with job start/arrival and end/departure records are easier if you reshape the data to have one data record recording each state-change event, tied by a common key.
So if your data has columns for Date, ID, Event, User -- where Event can be either Arrive or Depart -- then most visualizations will be much easier. In this view, each user that has arrived and departed will get 2 records instead of just 1. And customers that have arrived but not yet departed will get 1 record.
You can leave your original data in its original form and create this reshaped view in a Tableau Data Source using either Tableau Prep or the DataSource page in Tableau Desktop. You'll need to use the Union operator. That way you can have another data source that points to the original format if that is useful also.
I am trying to create multi-line subplots such that each subplot represents a rating factor on which firms are rated, and each line represents a firm's change in rating over time. Note that I don't have the date field. I have a string field that says which quarter of the year the Mean of ratings belong to.
I'm new to tableau and I created something like this:
I want to create multiple subplots: one subplot for each of the 8 rating features in the above diagram, and the subplot has multiple lines representing the variation over time for each company. I have 10 companies.
Please guide me on how to achieve this?
A rough sketch of how I Imagine the plot to look like is below:
This will teach you how to make a panel chart in Tableau: https://tarsolutions.co.uk/blog/build-a-dynamic-panel-chart-in-tableau/
It shouldn't to too complex for what you're attempting (fingers crossed!).
Building the panel chart is a little tricky in Tableau. Not terribly hard once you understand table calcs. I suggest starting simpler by moving Measure Names from the Rows shelf to the Pages shelf, then creating the line charts you want - 1 per page. Then learn about table calc partitions and addressing, including “At the Level” to have the info needed to define the panels
To make a colored line chart,
You need one field on columns that you set to display as continuous, making it green. If it is a date field, right click on the field and choose from the second set of groups, like month or quarter.
You need a continuous field on the Rows shelf, use Measure Values in your case
Use a discrete field on color to determine the number of lines, say Shortname
Set the Mark Type to either Automatic or line
As mentioned in camelot, we can extract table from particular region like:
tables = camelot.read_pdf('table_regions.pdf', table_regions=['170,370,560,270'])
But how can I find these regions for my pdf.
You can detect this regions, by some visual debugging.
https://camelot-py.readthedocs.io/en/master/user/advanced.html#visual-debugging
I know it's a late reply - but I just came across a possible solution.
If you're looking for a automated extraction method, you could use lattice in a first step, retrieve the table boundaries with tables[0]._bbox and use these numbers in a second call to camelot.read_pdf() into the argument table_areas.
Be aware that they are in a weirdly sorted format for a bbox.
If you just want to detect the table region you are reading, try to do this using Jupyter Notebook:
Define the table region inside .read_pdf method: tables = camelot.read_pdf('table_regions.pdf', table_regions=['170,370,560,270'], flavor='lattice'); pay attention on the flavor, because it defines whether the table have borderlines or not(it can be lattice for borders or stream for space).
Use camelot-py with plot from matplotlib: camelot.plot(tables[index], kind='contour') (You may know about how many index your object have by simply executing the name of the object. e.g.: tables runnign inside .ipynb cell)(contour is a visual debugging).
The plot will show an image of your table with a red rectangle contour. Just repeats step 2 until you achieve the table region you want to extract.
To test if the data is correct just use tables[index].df.
I have a group of trellis graphs on some data, in there you can see a numeric variable on the Y axis and a series of cell dishes on the X axis. Not all the numeric values are present on all the series of cells. Because of this the visualization results in a graph with empty spaces:
This is OK most of the time but the thing is I would like to avoid the "empty spaces, only in these graph series, that you can see between the bars. I would like to see showing only the pattern of the cell dishes where I have data.
Trying to do so I tried creating a calculated column to use it as a ordering index (https://docs.tibco.com/pub/sfire-bauthor/7.9.0/doc/html/en-US/GUID-8CAA18D0-CF28-4707-9945-041BDFD99E99.html) (Sorting Filter values asc/desc on Tibco Spotfire), after that "Limit data by expression" using a "[MyColumn] is not null" on it (https://community.tibco.com/questions/can-i-automatically-make-spotfire-ignore-empty-values-categories-charts) (How to show the top 10 column values in Spotfire) with no luck and I tried also (https://docs.tibco.com/pub/spotfire/6.5.1/doc/html/ncfe/ncfe_details_on_custom_expression.htm) create a custom expression, which I think it would be a good solution because I understand it will only affect these graphs and not the complete set of visualizations but no, I don't reach the point to change it. Last that it should work but it doesn't is to "Show/Hide Items" under a Boolean expression that it would include that "[Axis.Value] is not NULL" and "Apply individually for each trellis panel" of the numeric column which sound terrific but... nope, it didn't work either...
Any help would be appreciated, now I will select one by one on individual graphs extracting them and plotting them in other place but this is not very useful as a "large scale" solution. I am sure there is a way to insert a proper expression to avoid the null values of the cross of both variables, the numeric and the cell dishes.
this is because you are trellising data, not the axis. you won't be able to filter out values on the x axis; it's simply not how trellis works.
using multiple visualizations is the solution, but I assume you've got n sets of categories that you want to separate out without creating a ton of charts on the page and perhaps you can't guarantee the number of categories or their names, so you want to build a flexible solution.
please check out an answer I just wrote over here which illustrates how to use a document property and a property control to limit a visualization. your property control can be linked to automatically and dynamically display unique values in your "category" column (the one you are trellising by). maybe this can be a solution for you?
Noob here, I have a table with different entries (rows) per different (repeating) regions.
I'd like to be able to display the data rows filtered - matching that particular region thanks - so I get those particular fields related to each region as a tooltip on a map. (I know how to build the map)
Thank you
Just dragging the fields you want to Details or Tooltip is not doing the trick?
Putting a measure on a shelf (other than filter shelf) includes that field in the visualization query results -- i.e. applies the chosen aggregation function to yield an aggregate result value for each partition of the data (as specified by the unique combination of dimensions)
Putting a dimension on a shelf (other than the filter shelf) also includes that field in the query results, but since the dimensions define how data rows are partitioned, it can affect the level of detail of the query. You'll notice this often as suddenly getting many more marks in your visualization after you add a dimension to a shelf. If you are familiar with SQL, dimensions define the fields that follow the GROUP BY keyword.
EDIT
Thanks for the addition, #AlexBlakemore. I've never said dragging a dimension would not work, only that it wouldn't work as the OP wanted it to (basically the same as you're saying).
And though everything you said (above) is true, it's particularly not exact when it comes to maps. Yes, dragging a dimension will further partition the data, but it will not create additional marks on a map (unless it has also geographical properties). Rather, the tooltip will get the first occurrence of that dimension, and display data for that only. For instance, if you drag "Product" to details, and the possible values are "Bread", "Coffee" and "Milk", it will probably just show "Product: Bread", and the measures for "Bread" only. So yes, it will partition, but no, it won't create additional marks.
Back to OP problem. What I believe you want is a tool tip with all values of the dimension (in my poor example you'd like to see "Bread, Coffee, Milk"). Tableau does not have functions to aggregate strings yet, so it's hard to do so.
What I would suggest is to create a separate sheet, and just drag the dimensions and measures you want to rows. Then put it side by side with the map on a dashboard, and use the map as a filter. Then, when you click on a country/region/city, you'll see the data of that region on the other chart.
Refer to: http://kb.tableausoftware.com/articles/knowledgebase/creating-filter-actions-dashboards
or https://www.tableausoftware.com/learn/tutorials/on-demand/authoring-interactivity