Org-mode and gnuplot to track weight

weight_tracking.org

Tracking Weight

In addition to showing weight development, the following serves as documentation on how to use orgplot and the spreadsheet module.

Date	Observation	Trend	Change
[2017-09-10 Dom]	85.1	85.100	0.000
[2017-09-11 Seg]	84.9	85.080	-0.020
[2017-09-12 Ter]	83.1	84.882	-0.198
[2017-09-13 Qua]	82.6	84.654	-0.228
[2017-09-14 Qui]	83.7	84.559	-0.095
[2017-09-15 Sex]	83.9	84.493	-0.066
[2017-09-16 Sáb]	83.7	84.414	-0.079
[2017-09-17 Dom]	83.9	84.363	-0.051
[2017-09-18 Seg]	83.5	84.277	-0.086
[2017-09-19 Tue]	83.4	84.189	-0.088
[2017-09-20 Wed]	82.5	84.020	-0.169
[2017-09-21 Thu]	82.3	83.848	-0.172
[2017-09-22 Fri]	81.8	83.643	-0.205
[2017-09-23 Sat]	82.8	83.559	-0.084
[2017-09-24 Sun]	82.3	83.433	-0.126
[2017-09-25 Mon]	82.5	83.340	-0.093
[2017-09-26 Tue]	82.9	83.296	-0.044
[2017-09-27 Wed]	82.6	83.226	-0.070
[2017-09-28 Thu]	81.4	83.043	-0.183
[2017-09-29 Fri]	81.7	82.909	-0.134
[2017-09-30 Sat]	82.0	82.818	-0.091
[2017-10-01 Sun]	82.3	82.766	-0.052
[2017-10-02 Mon]	82.4	82.729	-0.037
[2017-10-03 Tue]	81.7	82.626	-0.103
[2017-10-04 Wed]	81.3	82.493	-0.133
[2017-10-05 Thu]	81.7	82.414	-0.079
[2017-10-06 Fri]	82.3	82.403	-0.011
[2017-10-07 Sat]	81.3	82.293	-0.110
[2017-10-08 Sun]	81.8	82.244	-0.049
[2017-10-09 Mon]	82.3	82.250	0.006
[2017-10-10 Tue]	81.2	82.145	-0.105
[2017-10-11 Wed]	80.5	81.981	-0.164
[2017-10-12 Thu]	79.7	81.753	-0.228
[2017-10-13 Fri]	80.2	81.598	-0.155
[2017-10-14 Sat]	81.0	81.538	-0.060
[2017-10-15 Sun]	81.5	81.534	-0.004
[2017-10-16 Mon]	81.3	81.511	-0.023
[2017-10-17 Tue]	80.8	81.440	-0.071
[2017-10-18 Wed]	79.9	81.286	-0.154
[2017-10-19 Thu]	79.8	81.137	-0.149
[2017-10-20 Fri]	80.1	81.033	-0.104
[2017-10-23 Mon]	81.3	81.060	0.027
[2017-10-24 Tue]	80.7	81.024	-0.036
[2017-10-25 Wed]	80.1	80.932	-0.092
[2017-10-26 Thu]	80.1	80.849	-0.083
[2017-10-27 Fri]	79.8	80.744	-0.105
[2017-10-30 Mon]	82.5	80.920	0.176
[2017-10-31 Tue]	81.4	80.968	0.048
[2017-11-01 Wed]	80.8	80.951	-0.017
[2017-11-02 Thu]	81.4	80.996	0.045
[2017-11-07 Tue]	79.9	80.886	-0.110
[2017-11-08 Wed]	79.4	80.737	-0.149
[2017-11-09 Thu]	80.3	80.693	-0.044
[2017-11-10 Fri]	80.5	80.674	-0.019
[2017-11-11 Sat]	81.0	80.707	0.033
[2017-11-13 Mon]	81.2	80.756	0.049
[2017-11-14 Tue]	80.2	80.700	-0.056
[2017-11-15 Wed]	80.5	80.680	-0.020
[2017-11-16 Thu]	79.7	80.582	-0.098
[2017-11-17 Fri]	79.4	80.464	-0.118
[2017-12-04 Mon]	81.6	80.578	0.114
[2017-12-06 Wed]	80.5	80.570	-0.008
[2017-12-11 Mon]	81.0	80.613	0.043
[2017-12-12 Tue]	80.8	80.632	0.019
[2017-12-14 Thu]	80.1	80.579	-0.053
[2018-01-02 Tue]	82.5	80.771	0.192
[2018-01-03 Wed]	81.4	80.834	0.063
[2018-01-05 Fri]	81.3	80.881	0.047
[2018-01-06 Sat]	81.0	80.893	0.012
[2018-01-08 Mon]	81.7	80.974	0.081
[2018-01-10 Wed]	81.1	80.987	0.013
[2018-01-11 Thu]	81.4	81.028	0.041
[2018-01-13 Sat]	81.5	81.075	0.047
[2018-01-14 Sun]	82.2	81.188	0.113
[2018-01-16 Tue]	81.7	81.239	0.051
[2018-01-19 Fri]	80.5	81.165	-0.074
[2018-01-23 Tue]	81.1	81.159	-0.006
[2018-01-24 Wed]	81.0	81.143	-0.016
[2018-01-25 Thu]	80.7	81.099	-0.044
[2018-01-26 Fri]	80.3	81.019	-0.080
[2018-01-29 Mon]	82.1	81.127	0.108
[2018-02-01 Thu]	79.1	80.924	-0.203
[2018-02-02 Fri]	78.9	80.722	-0.202
[2018-02-03 Sat]	80.1	80.660	-0.062
[2018-02-04 Sun]	80.5	80.644	-0.016
[2018-02-05 Mon]	80.9	80.670	0.026
[2018-02-06 Tue]	80.6	80.663	-0.007
[2018-02-12 Mon]	80.4	80.637	-0.026
[2018-02-13 Tue]	79.9	80.563	-0.074
[2018-02-14 Wed]	79.9	80.497	-0.066
[2018-02-15 Thu]	79.7	80.417	-0.080
[2018-02-16 Fri]	78.1	80.185	-0.232

Now, this shows an interesting thing. When a table is sent to gnuplot via Babel, the dates are formatted like 2017-09-13-00:00:00, so for gnuplot to interpret the x range correctly, we must let gnuplot know that we are dealing with time data, and explicitly declare the format.

Struggled with the date format for a while until I found this post: https://lists.gnu.org/archive/html/emacs-orgmode/2012-03/msg01019.html

To get vertical lines from trend to observation, I plot with yerrorbars. The 4 values are X, Y, minY maxY, so we get an observation at Y, and a line from the observation to the trend.

My latest discovery is that gnuplot can produce svg directly, by setting the terminal to svg and outputting the results as shown below:

reset
set terminal svg size 800,600
set title "Weight tracking - Helge"

set xlabel "Date of measurement"
set xdata time
set timefmt "%Y-%m-%d-%H:%M:%S"

set format x "%Y-%m-%d"
set xtics nomirror rotate by -45

set autorange x
set autorange y

set ylabel "Weight in kg"

unset bars
set rmargin 10
plot data u 1:3 smooth csplines lw 2 title 'Trend', \
     data u 1:2:2:3 w yerrorbars lw 1 pt 7 title 'Observation'

Formula explanation: To avoid referencing values outside the table range, check for current row in formula https://emacs.stackexchange.com/questions/32279/org-mode-add-formula-to-generate-delta-column