csv2graph

This script generates a PDF graph from a CSV file.

Description

This script generates a PDF graph from a CSV file. It has two modes:

1. line graph (default)
2. pie chart

By default, the input is read from stdin and printed to stdout.
You can also specify files with the --input and --output option.

When you use the --output option you can specify a different output format from PDF (e.g. png).

line graph

Generate a line graph from the data.
The first value of each line is used as a descriptive Label for the following data. The first line must consist of dates according to the format specified in --date-format option (default: %Y-%m). The dates can contain a label. You can state the separator between the date and the label via --x-label-separator. The date can contain the seperator. The script will find the right spot to split.

pie chart

Generate a pie chart from the data.

installation

You need to install these required packages: numpy, scipy, matplotlib
Or use the requirements.txt file:

pip install -r requirements.txt

usage & options

usage line: csv2graph [-h] [--input FILE] [--output FILE]
                      [--start-at-zero] [--threshold T] [--smooth] [--stacked] [--disable-markers]
                      [--data-in-columns] [--title TITLE] [--emphasize LABEL [LABEL ...]]
                      [--date-format FORMAT] [--x-label-separator SEPARATOR]
                      [--annotations FILE]
                      [--second-y-axis [AXIS]]

usage pie: csv2graph --pie-chart [-h] [--input FILE] [--output FILE]
                     [--data-in-columns] [--title TITLE]

optional arguments:
  -h, --help            show this help message and exit
  --smooth, -S          smooth data
  --stacked, -s         stacks data on top of each other. (Don't use with --second-y-axis).
  --start-at-zero, -z   force y-axis to start at zero
  --threshold T, -T T   print a threshold in the chart
  --pie-chart, -p       create a pie chart
  --input FILE, -i FILE
                        read data from this file instead of stdin
  --output FILE, -o FILE
                        print chart to this file instead of stdout
  --data-in-columns, -c
                        data is stored column-wise not row-wise
  --date-format FORMAT, -d FORMAT
                        set format for dates (default: %Y-%m)
  --x-label-separator SEPARATOR, --separator SEPARATOR
                        set separator between date and label
  --title TITLE, -t TITLE
                        set title
  --emphasize LABEL [LABEL ...], -e LABEL [LABEL ...]
                        emphasize Label by printing the line wider
  --annotations FILE
                        add annotations from file FILE
  --second-y-axis [AXIS]
                        Add second y-axis with different scaling.
                        Specify the axis for the dataset as a sequence of '1' or '2'.
                        (Don't use with --stacked.)
  --disable-markers, --no-markers, -m
                        disable markers on datapoints                  

examples

line

line.csv:

month,2019-12,2020-01,2020-02,2020-03,2020-04,2020-05,2020-06,2020-07,2020-08,2020-09,2020-10,2020-11
"own issues",48,40,68,67,47,29,51,49,55,46,100,93
"foreign issues",15,18,17,23,13,11,38,30,11,5,27,18
other,109,113,124,178,209,139,142,135,148,160,150,148

Command:

./csv2graph --title "Issues per Month" -i examples/line.csv -o examples/line.pdf --date-format "%Y-%m" --emphasize "own issues"

output png

Command:

./csv2graph --title "Issues per Month" -i examples/line.csv -o examples/line.png --date-format "%Y-%m"

Disable markers

Command:

./csv2graph --title "Issues per Month" -i examples/line.csv -o examples/line_no_markers.png --date-format "%Y-%m" --disable-markers

data in columns

If your data is column wise, you have to add the --data-in-columns option

line_columns.csv:

month,own issues,foreign issues,other
2019-12,48,15,109
2020-01,40,18,113
2020-02,68,17,124
2020-03,67,23,178
2020-04,47,13,209
2020-05,29,11,139
2020-06,51,38,142
2020-07,49,30,135
2020-08,55,11,148
2020-09,46,5,160
2020-10,100,27,150
2020-11,93,18,148

Command:

./csv2graph --title "Issues per Month" --data-in-columns -i examples/line_columns.csv -o examples/line_columns.pdf --date-format "%Y-%m" --emphasize "own issues"

Stacked data

You can stack the data.

Command:

./csv2graph --title "Issues per Month" -i examples/line.csv -o examples/line-stacked.pdf --date-format "%Y-%m" --stacked --smooth

line with labels

line_with_label.csv:

month,2019-12-dec,2020-01-jan,2020-02-feb,2020-03-mar,2020-04-apr,2020-05-may,2020-06-jun,2020-07-jul,2020-08-aug,2020-09-sep,2020-10-oct,2020-11-nov
"own issues",48,40,68,67,47,29,51,49,55,46,100,93
"foreign issues",15,18,17,23,13,11,38,30,11,5,27,18
other,109,113,124,178,209,139,142,135,148,160,150,148

Command:

./csv2graph --title "Issues per Month" -i examples/line_with_label.csv -o examples/line_with_label.pdf --date-format "%Y-%m" --x-label-separator "-"

line with annotations

You can add annotations to your graph.

In addition to the data csv file, you need a csv file with the annotations.

The text can contain newlines.

annotations.csv:

2020-02,2020-09
"release AI feature","remove AI feature"

Command:

./csv2graph --title "Issues per Month" -i examples/line.csv -o examples/line_annotations.pdf --date-format "%Y-%m" --annotations examples/annotations.csv

data in columns

If your data is column wise, the annotations also have to be column wise

annotations_columns.csv:

2020-02,"release AI feature"
2020-09,"remove AI feature"

Command:

./csv2graph --title "Issues per Month" --data-in-columns -i examples/line_columns.csv -o examples/line_annotations_columns.pdf --date-format "%Y-%m" --annotations examples/annotations_columns.csv

overlapping text

If your text overlaps, you can define an offset.

The first offset applies in x direction in the scale of the date. (The dates are converted to a float representing dates in the unix epoche, see matplotlib.date2num). The second offset applies in y direction in the scale of the data.

annotations_overlap.csv:

2020-02,2020-03,2020-09,
"release AI feature","This text 
will overlap","remove AI feature"

Command:

./csv2graph --title "Issues per Month" -i examples/line.csv -o examples/line_annotations_overlap.pdf --date-format "%Y-%m" --annotations examples/annotations_overlap.csv

annotations_offset.csv:

2020-02,2020-03,2020-05,2020-07,2020-09
"release AI feature (offset x and y)","This text
will overlap","offset x","offset y","remove AI feature"
-20,,10,,
5,,,20,

Command:

./csv2graph --title "Issues per Month" -i examples/line.csv -o examples/line_annotations_offset.pdf --date-format "%Y-%m" --annotations examples/annotations_offset.csv

2 different y axis

You can use 2 different y-axis with different scaling.

If there are more than 2 datasets, you have to specify to which axis the dataset belongs, with a sequence of 1 or 2.

You can't use this together with --stacked.

Command:

./csv2graph --title "Issues per Month" -i examples/line.csv -o examples/line-second-y-axis.pdf --date-format "%Y-%m" --second-y-axis 112

pie

input-file:

"ips",64
"dns",22
"frontend",19
issue-1234,8
issue-9874,5
other,62

Command:

./csv2graph --title "Most time consuming issues" -i examples/pie.csv -o examples/pie.pdf --pie

Dependencies and Tests

The tests generate the examples and compare the results to the existing ones. Unfortunately, the output depends on the matplotlib version. That is why we pin all dependencies to a specific version using pip-tools.