python script that converts CSV files into SVG images
Requires python's builtin csv module and the pygal module.
Turns one or more CSV files into SVG image charts. Can concatenate multiple CSV files into one logical database, from which it can chart multiple values as different charted lines. Exposes a fair amount of the PyGal functionality. although documentation of such is terrible.
Example 1 Chart temperature and humidity from one logical data base, contained in five CSV files.
The start of log1.csv looks like this:
date,time,tempC,tempF,pressurePA,presureMB,light,R,G,B,heading,magnetometer0,magnetometer1,magnetometer2,accelX,accelY,accelZ,anal0,anal1,anal2,anal3,leds
2018-03-11,18:18,26.91,80.44,98553.97,985.54,61,104,112,125,336.97,2089,-1203,-5017,-0.01,0.05,1.0,0.561,0.543,0.549,0.585,OFF
2018-03-11,18:35,27.12,80.81,98580.4,985.8,245,106,83,94,336.57,2098,-1197,-4989,-0.01,0.05,1.0,0.516,0.57,0.582,0.555,ON
2018-03-11,19:05,26.77,80.18,98598.58,985.99,396,165,133,137,335.08,2100,-1212,-4998,-0.0,0.05,0.99,0.549,0.558,0.537,0.612,ON
log2.csv, log3.csv, etc have the same fields, and all files have disjoint data (no overlaps--they represent five days worth of readings).
We can create a chart of the temperature using this command:
csvchart.py \
"csv:file=log1.csv,log2.csv,log3.csv,log4.csv,log5.csv;x-field=date,time;y-fields=tempC;labels=Celcius" \
"chart:file=example1a.svg;title=Example 1a temperature readings;interpolation=cubic;"
Note the structure of this command, The csvchart.py command takes two (or more) arguments. The first is a "csv:..." argument, and it specifies the data source or sources. The rest of this argument are key/value pairs seperated by semicolons. The values may be lists, which are separted by commas. There's plenty to say here, but the most salient comments are that the file= key accepts a list of CSV files that needs to have disjoint data ranges. The x-field= will be interepretted as the single X-axis variable, and the value(s) come from the CSV file. It can be multiple values, which will be concatenated together before being added to the chart. This allows you to use CSV records where the date and time are seperate fields. Whereas the y-fields values can also be a list, but each value will be charted separately (e.g. temperature vs humidity). If you have multiple y-fields, you can have multiple labels, just keep then in the same order.
The second argument is the "chart:..." argument. It specifies how the resulting SVG chart looks. In this argument, most of the values are singletons, not lists. For example, the file= key/value will specify the single SVG output file. Note that the chosen interpolation can be very misleading. Right now, csvchart lets you choose any of the interpolation systems provided, but does not yet expose further attributes like interpolation_parameters or interpolation_precision.
The result of the above command on the example 1 data looks like 
.
Here's a different version of the above command. It will display an average of each day's temperatures, and will also draw in the min and max temperatures (drawn as confidence intervals )
csvchart.py \
"csv:file=log1.csv,log2.csv,log3.csv,log4.csv,log5.csv;x-field=date;y-fields=tempC;labels=Celcius" \
"chart:file=example1b.svg;title=Example 1b average daily temperature;interpolation=cubic;value=confidence"
Which results in 
.
Note that instead of x-field=date,time, this new command only has x-field=date in the "csv:..." argument. Since there are multiple records with the same date, by default, only the last record of that day would get charted. But by adding value=confidence to the "chart:..." argument, all the values for that day will be averaged, and then confidence intervals based on the minimum and maximum for that day. Other value= values are min, max, mean, first, and last. A value= of maxconfidence graphs the maximum value for the day, while still showing the min/max confidence interval. LIkeways minconfidence.
If this is confusing, consider that the behaviour of value= in the "chart:..." argument has an aggregation effect that's similar to GROUP BY in SQL SELECT statements. No doubt you are no longer confused.
Example 2 Chart temperature and humidity from three logical data bases, contained across fifteen different CSV files.
TODO