Rhythm

Ups and Downs, Rhymes and Reasons, Tides and Ebbs, Seasons and Rhythms

What graphs will we see today?

Variable #1	Variable #2	Chart Names	Chart Shape
Quant	Quant	Line Plot

What kind of Data Variables will we choose?

No	Pronoun	Answer	Variable/Scale	Example	What Operations?
1	How Many / Much / Heavy? Few? Seldom? Often? When?	Quantities, with Scale and a Zero Value.Differences and Ratios /Products are meaningful.	Quantitative/Ratio	Length,Height,Temperature in Kelvin,Activity,Dose Amount,Reaction Rate,Flow Rate,Concentration,Pulse,Survival Rate	Correlation

Inspiration

(a) Line Plot Inspiration https://fivethirtyeight.com/features/katie-ledecky-is-the-present-and-the-future-of-swimming/

(b) Time Series Plot Inspiration

Figure 1: Line Plots

Ek Ledecky bheegi-bhaagi si, is it? Yeh Ledecky hai, ya jal-pari?

In Figure 2 (a), the black line is the average of the 50 best times at each distance since 2000. The top 200 times for each distance since 2000 are also plotted, with light orange lines each representing one swimmer. Her races and her career essentially follow the same pattern — the more she swims, the more she separates from the field. Her 1500 metres record timing is better than the best time for 800m!!😱 (Update July 2024: Ledecky won the bronze at Paris 2024)

And LA? The weather in California…ahh. But Seattle has more variation, and sudden changes too!

NoteWhy are fewer babies born on weekends? And more in September?

(a) Births over Days of the Week

(b) Births over Months of the Year

Figure 2: Solomon Grundy, born on Monday… was an Accountant?

Looks like an interesting story here…there are significantly fewer births on average on Sat and Sun, over the years! Why? Should we watch Grey’s Anatomy ?

And why more births in September? That should be a no-brainer!! 😂

How do these Chart(s) Work?

Line Plots take two separate Quant variables as inputs. Each of the variables is mapped to a position, or coordinate: one for the X-axis, and the other for the Y-axis. Each pair of observations from the two Quant variables ( which would be in one row!) give us a point. All this much is identical with the Scatter Plot.

And here, the points are connected together and sometimes thrown away altogether, leaving just the line.

Looking at the lines, we get a very function-al sense of the variation: is it upward or downward? Is it linear or nonlinear? Is it periodic or seasonal…all these questions can be answered with Line Charts.

ImportantWhen one variable is Time?

Line charts often have one variable as a time variable. In such case the data is said to be a time series.

Any metric that is measured over regular time intervals forms a time series. Analysis of Time Series is commercially important because of industrial need and relevance, especially with respect to Forecasting (Weather data, sports scores, population growth figures, stock prices, demand, sales, supply…). For example, in the graph shown below are the temperatures over time in two US cities:

Plotting a Line Plot

Using Orange

Let us at least look at this data in Orange, and import it into this rather elaborate Orange Workflow:

Import this into Orange and see…

Figure 3: RIAA Revenues Data Table

Figure 3 states that there are 432 data points, with 7 variables in the dataset; some missing data.

For now, the variables we need are :

NoteQuantitative Data

• Year: (int) Year in which RIAA revenue was logged
• Value (For Charting): (int) Revenue in million USD We can ignore the rest for now, unless we plan to work more with this data, and need to know more. The other numerical data showing billions of USD are not easily decipherable, an example of data that is not documented well…

NoteQualitative Data

• Category: (chr) Form of the Music released ( CD etc..)

We need to first form time series from the dataset: we will choose the year-date variable, and indicate that it starts on Jan 1, 1973:

(a) Forming Time Series

(b) Plotting Time Series

Figure 4: Forming and Plotting Time Series in Orange

Using RAWgraphs

Upload this RAWgraphs project tutorial file into https://app.rawgraphs.io/ and play! Here is something we can create there:

Using DataWrapper

https://academy.datawrapper.de/article/23-how-to-create-a-line-chart

Here be dragons: DataWrapper wants the data in wide format: each Format of music needs to have its figures in a separate column! 🤦. And this is not a data transformation that we can achieve within DataWrapper. Bah.

We are probably better off plotting a regular scatter plot. Here too there seem to be limitations because we are not able to colour the series based on type of music Format.

The Shape of You Data

Never mind that silly song now.

As mentioned above, data can be in wide or long form. How does one imagine this shape-shifting that seems needed now and then? Let’s see.

ImportantLong Form and Wide Form Data

Several tools such as DataWrapper (and others, yes, I agree, even with code, as we will see) need data transformed to a specific shape. this is usually mandated by the “shape or geometry” we intend to use in the visualization. We should now look at this idea of shape in data. Consider the data tables below:

Product	Power	Cost	Harmony	Style	Size	Manufacturability	Durability	Universality
G1	0.5858003	0.2773750	0.7244059	0.0731445	0.1000535	0.4551024	0.9622046	0.9966129
G2	0.0089458	0.8135742	0.9060922	0.7546750	0.9540688	0.9710557	0.7617024	0.5062709
G3	0.2937396	0.2604278	0.9490402	0.2860006	0.4156071	0.5839880	0.7145085	0.4899432

Figure 5

Product	Parameter	Rating
G1	Power	0.5858003
G1	Cost	0.2773750
G1	Harmony	0.7244059
G1	Style	0.0731445
G1	Size	0.1000535
G1	Manufacturability	0.4551024
G1	Durability	0.9622046
G1	Universality	0.9966129
G2	Power	0.0089458
G2	Cost	0.8135742
G2	Harmony	0.9060922
G2	Style	0.7546750
G2	Size	0.9540688
G2	Manufacturability	0.9710557
G2	Durability	0.7617024
G2	Universality	0.5062709
G3	Power	0.2937396
G3	Cost	0.2604278
G3	Harmony	0.9490402
G3	Style	0.2860006
G3	Size	0.4156071
G3	Manufacturability	0.5839880
G3	Durability	0.7145085
G3	Universality	0.4899432

Figure 6

What we have done is:

• convert all the variable names into a stacked column Parameter
  
• Put all the numbers into another column Rating
  
• Repeated the Product column values as many times as needed to cover all Parameters (8 times).
  

See the gif below to get an idea of how this transformation can be worked reversibly. (Yeah, never mind the code also.)

Figure 7: Data Pivoting

Figure 8: Data Stacking in RAWgraphs

So how can we actually do this? Two Ways.

Turns out there are some nice people at U. San Diego who have built an R-oriented app called Radiant for Business Analytics that can do this pretty much click-and-point style, though it is nowhere as much fun as Orange. Head off there:

https://vnijs.shinyapps.io/radiant

We upload our original data, pivot it, and download the pivotted data. Now the pivotted wide-form data should work in DataWrapper.

And RAWgraphs also has a stack on column option that does pretty much the same thing. See here: https://www.rawgraphs.io/learning/how-to-stack-your-unstacked-data-or-meet-the-unpivoter

Whatever, peasants.

What is the Story here?

• Over the years different music formats have had their place in the sun
• All physical forms are on the wane; streaming music is the current mode of music consumption.

Dataset: Weather at New York City Airports

To get an idea of seasons, trends and to try our hand at time-series forecasting, let us look at a data set pertaining to the weather at New York city airports.

Examine the Data

Included below is a PDF report from Orange, summarizing the data, generated from the Feature Summary widget::

Download PDF File

Figure 9: Weather Data Table

We should take the first column time_hour and see if we can use that as our time variable. All the weather related numerical data columns are individual time series which we can plot and analyse.

Data Dictionary

NoteQuantitative Data

• time_hour(num): Numeric date-time variable. Does Orange spot this?
• year(num): Just 2013.
• month, day, hour(num): components of the exact time of measurement of weather parameters
• humid,temp,wind_dir, wind_speed, wind_gust, precip, pressure, visib (num): all numeric weather parameters

NoteQualitative Data

• origin (text): airport (JKF/EWR/LGR)

Let us build an Orange workflow step-by-step for this dataset and its Research Questions.

Research Questions

There are a lot of parameters to play with and investigate here.

NoteQuestion

Q1. What is Temperature temp over time at each of three airports?

This is a Scatter Plot of course.

It seems the Line Chart widget in Orange cannot colour individual time series by colour using another Qualitative variable. 😢. Is there a better way? (You know the answer.)

Also note the utter busy-ness of this chart. This is a chart of 26K points, well beyond what we can digest at one time. We need to summarize/average etc.

NoteQuestion

Q2. In the US, there is a lot of talk of “wind chill factor”. So, is there some graphical evidence of windchill? (temp and wind_speed)?

Figure 10: NYC Airport Temperatures over Time

NoteQuestion

Q3. How do averaged plots look like, for temp, humid, and dewpoint?

We can use the Moving Transform widget in Orange to calculate monthly averages for these quantities, after converting the data into a time series.

Figure 11: Time Averaged Weather Plots-3

What is the Story Here?

• There is a strong natural seasonal trend over the period of one year in the temperature at all three airports
• If we plot temperature against windspeed, we see a fair negative slope/correlation, as we would expect.
• Humidity is high most times, except during some very dry winter months?

Note

Did you notice the serious outlier in the temp vs windspeed graph? Try to remove the Select Rows widget and see if you can spot it. Do you understand why that egregious reading had to be be filtered?
Such readings are called outliers.

---

Dataset: Born in the USA

Tourist: Any famous people born around here?
Guide: No sir, best we can do is babies.

The Time Series Line Chart widget in Orange is described here. https://orangedatamining.com/widget-catalog/time-series/line_chart/

Let us take some Births related data and plot it in Orange.

And download the Line Chart workflow file for this data:

Note how we have two widgets for the Line Charts. More shortly.

Examine the Data

Figure 12: Born in the USA

Figure 13: Births Summary Table

Data Dictionary

NoteQuantitative Data

• year, month, date_of_month: (int) Columns giving time information
• day-of_week: (int) Additional Time information
• births: (int) Total live births across the USA that day

NoteQualitative Data

None. Though we might covert day_of_week and month into Qual variables later.

Evenly spread year, month, date_of_month and day_of_week variables…the bumps are curious though, no? day_of_week is of course neat. births are numerical data and have a good spread with a bimodal distribution distribution. Some numbers in the mid-range hardly occur at all… So a premonition of some two-valued phenomenon here already.

Research Questions

NoteQ1. What does the births data look like over the years?

Figure 14: Births over the Years

Hmmm…very busy graph. The overall trend is a slight bump in births around 2007 and then a slow reduction in births. Large variations otherwise, which we need to see in finer detail on a magnified scale, a folded scale, or by averaging.

Converting month or day_of_week to categorical in the File Menu does not provide us with a way of separating the time series by month or weekday…sad.. We will be able to average over month, day_of_week to see what happens.

NoteQ2. What do births look like averaged over month?

This is good! We have converted the dataset to a timeseries, of course, and then added a moving transform widget, that allows us to take averages of births over weeks, months, or years. Play with this setting in the moving transform widget.

We see that averaging i.e. $aggregating$ by Month of year clearly shows September as the month for the most number of births.

NoteQ3. What do births look like averaged over day_of_week?

Here too with the moving transform widget, choosing Day of Week as the aggregating parameter, we see a dip in births over weekends. Try!!

ImportantFolded Scale?

Look at the figure below.

Figure 15: Aggregate over Week of the Year

It should be apparent that the line chart shows averages based on “Week of Year”. What does that mean?

Imagine a carpenter’s folding footruler:

Imagine the entire time series stretched out and then folded over itself at intervals of a week. There will of course be overlapping data that represent data points for the same week year after year. THAT is what goes into the averaging!

So we see that the weeks in September show the highest average birth numbers, which seems right!

Other Plots

Imagine that we follow this overlap routine and get the data by same-week-of-year, as before. We need not necessarily average that data; we can simply plot each (repeated) week’s worth of data as a box plot. This results in an array of boxplots, one per week, and is called a candlestick plot. Clearly we can do this for months, weeks, and even days of the week. Here is what it looks like; it does not seem possible to create these with any of the tools we are currently using.

Figure 16

As before, the medians are the black lines across each boxplot, which is one for each month. Note that since the medians are towards the upper end of the boxplots, we can guess that the per-month distribution must be skewed to the left (lower than median values are less frequent).

If the Quantities that vary over time are not continuous but discrete values such as high, medium, and low,, a time-series heatmap is also a possibility.

Figure 17

Very arbitrarily slicing the birth numbers into three bins titled high, fine, and low, we can plot a heatmap like this. Orange does have a heatmap widget, however it seems suited to Machine Learning methods such as Clustering.

Your Turn

NoteValentine’s Day Spending by Age

A regular line plot, not a time series.

https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2024/2024-02-13/gifts_age.csv

NoteWilliam Farr’s Data on Cholera in London, 1849

https://vincentarelbundock.github.io/Rdatasets/csv/HistData/Cholera.csv

NoteArctic and Antarctic Sea Ice coverage over time.

Is global warming affecting ice coverage at the poles?

This data is in wide form, and you may have to massage it into long form before pulling it into Orange!

NotePurple Air

In the Air Tonight: Head over to Purple Rain Purple Air and download air quality data from community based air quality sensors. Plot these as time series, and try getting historical data, or data on festivals or important occasions in specific cities.

Wait, But Why?

• Line Charts show up functional relationships or overall trends in the data.
• They can be made less cluttered than the corresponding scatter plots, especially with averaging.
• Seasonal cycles can also be spotted very easily.
• The X-axis need not necessarily be time: it can often be other (independent) variables, and the Y-axis plots the target/dependent variable.
• However, we do encounter many things that vary over time: weather, wealth, No. of users or downloads of an app, hits to a webpage, customers at a supermarket, or population of animals or plants in a region.
• These are best represented by Line Charts
• As humans, we are also deeply interested in patterns of recurrence over time, and in forecasting for the future, using tech, and using say Oracles.
• Both these purposes are amply served by Line Charts.

References

1. Charles Chambliss (1989). The Mundanity of Excellence: An ethnographical report on Stratification and Olympic Swimmers.
2. Nijs V (2023). radiant: Business Analytics using R and Shiny. R package version 1.6.0, https://github.com/radiant-rstats/radiant.
3. Robert Hyndman, Forecasting: Principles and Practice (Third Edition).available online
   
4. Time Series Analysis at Our Coding Club
   
5. The Nuclear Threat—The Shadow Peace, part 1
   
6. 11 Ways to Visualize Changes Over Time – A Guide