stemgraphic (@stemgraphic) header
stemgraphic (@stemgraphic) profile picture

Stemgraphic is a Python module and a command line tool that can be used to analyze distribution of data and outliers

North Carolina, USA
Joined June 2017
According to the Laboratory Handbook of Statistical Methods (1931), we do charts or tables. But in fact, it is not an exclusive or. We can do both at the same time. and there are also semi-graphical displays, such as stemgraphic.org (@stemgraphic) #dataviz #infoviz
1
2
1
2
Do you need to gain insight on your data? See the multimode? See the distribution shape? Potential issues? Check out this article about our free tool @stemgraphic blog.dionresearch.com/2019/1…
Did you know stemgraphic also has a wide range of semi-graphic displays you can use from the console, in a python or ipython session (or in a notebook even, see: github.com/dionresearch/stem…) #stats #python #infoviz #dataviz
2
3
1
1
Replying to @f_dion @stemgraphic
or using the digit glyph for other languages. For example in this case I used charset='arabic' on a stem-and-leaf plot (stem_text)
1
1
1
1
Replying to @f_dion @stemgraphic
The unicode charset support is also pretty handy to have, say, "bold" text. Even in a terminal. Simply add charset="bold" to the call (heatmap in this example with zipcodes).
1
2
Replying to @f_dion @tkasasagi
Tumblr was also degraded quite a bit. I've uploaded it to google photo: photos.app.goo.gl/d8QHQ5NBX1…
1
3
4
Replying to @stemgraphic
Probably one feature that few people know about, @stemgraphic provides a bunch of text and semi-graphic displays. From the command line or from ipython, or in the notebook, but as text. For example, good old tally chart
1
3
1
Of course with @stemgraphic, you can do stem-and-leaf plots in text mode, wide format now. Millions of rows in a few seconds on my laptop. And, if you’ve missed it, @stemgraphic has been able to do them with categorical or text too #dataviz #infoviz
4
3
Also, violin plots seem to suggest plays outside of the rink... Not the case with stem-and-leaf. Here's an example with NWHL data, 2017 season, using @stemgraphic to plot it (using pandas to select year and team, BOS), in a back to back format, for X coordinate:
1
1
That's a back to back @stemgraphic in case you are wondering. Distances are read directly on the stem and the fractional feet are the leafs. gives the *exact* distribution, unlike a box plot or violin plot.
In the above, I mean "distance of shots when there is a goal", to be clear. But what about looking at only the @WhitecapsHockey and @BuffaloBeauts games against each other? Buffalo is very consistent against any opponent and appears to block Minnesota's attack from distance.