Two Regression Lines in single plot, shaded by year, and points indicating a different location

JVanslinger25 · January 28, 2019, 7:32pm

When using ggplot2, is there a way in which I can plot two regression lines from the same data set? The two regressions would be categorized in the data set (P VS NP), however I want to indicate location category by point shape (each location represented by a different point shape), and year by shading (each year represented by a different color shade).

I can easily plot two regressions indicating P and NP, with colors showing each location. But I am not sure how I would show year.

andresrcs · January 28, 2019, 7:44pm

We don't really have enough info to help you out. Could you ask this with a minimal REPRoducible EXample (reprex)? A reprex makes it much easier for others to understand your issue and figure out how to help.

JVanslinger25 · January 28, 2019, 8:39pm

Year	Lake	Total Wt (g)	Total Length (mm)	K	Parasites
2015	A6	0.0915	29	0.375169134	0
2015	A6	0.166	34	0.42234887	0
2015	A6	0.201	36	0.430812757	0
2015	A6	0.1416	33	0.394022873	0
2015	A6	0.1913	35	0.446180758	0
2015	A8	0.365	41	0.529591852	0
2015	A8	0.534	48	0.482855903	0
2015	D7	0.182	35	0.424489796	0
2015	D7	0.1951	34	0.496387136	0
2015	D7	0.2137	34	0.543710564	0
2015	D7	0.2729	39	0.460054957	0
2015	D7	0.3845	44	0.451375845	0
2016	A6	0.0534	24	0.386284722	0
2016	A6	0.7558	48	0.683412905	0
2016	A6	0.6106	48	0.552119502	0
2016	A8	0.1219	29	0.499815491	0
2016	A8	0.1249	31	0.419254137	0
2016	A8	0.2519	37	0.497305194	0
2016	A8	0.4617	45	0.506666667	0
2016	A8	0.415	42	0.560144693	0
2016	A8	0.9104	51	0.686312203	0
2016	B2	0.0643	24	0.465133102	0
2016	B2	0.0807	27	0.409998476	0
2016	B2	0.1435	30	0.531481481	0
2016	B2	0.0292	19	0.425718035	0
2016	B2	0.0689	23	0.566285855	0
2016	B2	0.923	52	0.656434911	0
2016	D7	0.706	48	0.638382523	0
2016	D7	0.761	54	0.483285068	0
2016	D7	0.506	53	0.339877886	0
2016	D7	0.579	57	0.312646806	0
2016	D7	0.689	47	0.663629446	0
2016	D7	0.562	55	0.337791134	0
2016	D7	0.6049	51	0.456008624	0
2017	B2	0.264	35	0.61574344	0
2017	B2	0.3025	37	0.597200561	0
2017	B2	0.2666	36	0.571416324	0
2017	B2	0.3032	37.5	0.574957037	0
2017	B2	0.3573	41	0.51841964	0
2017	D7	0.6741	53	0.452789887	0
2017	D7	0.7112	49	0.604510026	0
2017	D7	0.7173	51	0.540742248	0
2017	D7	0.7104	50	0.56832	0
2017	D7	0.9858	59	0.479990651	0
2015	A6	0.5854	47	0.563844235	1
2015	A8	0.5298	48	0.47905816	1
2015	A8	0.8723	51	0.657590218	1
2015	D7	0.5171	46	0.531252568	1
2015	D7	0.55	47	0.529747744	1
2015	D7	0.659	46	0.677036246	1
2015	D7	0.5147	46	0.528786883	1
2016	A6	1.3292	61	0.585599676	1
2016	A6	0.3911	45	0.429190672	1
2016	A6	0.696	47	0.67037169	1
2016	A6	0.9489	55	0.570338092	1
2016	A6	0.6278	44	0.736992863	1
2016	A8	0.1397	34	0.35543456	1
2016	A8	0.2709	35	0.631836735	1
2016	A8	0.72	51	0.542777665	1
2016	A8	0.6535	49	0.555465835	1
2016	A8	0.8723	53	0.58591992	1
2016	A8	0.9278	57	0.500990858	1
2016	A8	1.1383	57	0.614656062	1
2016	B2	0.13	36	0.278635117	1
2016	B2	1.1902	57	0.642680879	1
2016	D7	0.2864	42	0.386567325	1
2016	D7	0.4656	41	0.675556071	1
2016	D7	0.3965	45	0.435116598	1
2016	D7	1.1652	58	0.597195457	1
2017	A6	1.7541	68	0.557863067	1
2017	B2	0.8721	54	0.553840878	1
2017	D7	0.5807	55	0.349030804	1

Ok so here is my reprex data set.

The code I have tried is:

Plot1<- ggplot(YearsPooled, aes(x=LogLength, y=LogWeight, colour=Year, linetype=Parasites)) + geom_point(size=2) + stat_smooth(method = "lm", col = "black", size=0.25, se=FALSE, fullrange=TRUE) + theme_bw() + theme(plot.background = element_blank(), panel.grid.minor = element_blank(), panel.border = element_blank()) + theme(axis.line = element_line(color = 'black')) + guides(shape=guide_legend(title=NULL), linetype=guide_legend(title=NULL)) + scale_color_grey("Year") + xlab("Length ln(mm)") + ylab("Weight ln(g)") + scale_shape_manual("Lake", values=c(15,16, 17, 18))

And I am getting this figure:

What I am trying to do and figure out is to change each point to represent the lakes in the data set by shape, while maintaining the regressions of Parasitized vs non parasitized fish, also I want to keep the color shading by year. Hopefully this makes more sense now. Thank you for quickly replying as well.

andresrcs · January 28, 2019, 9:18pm

Your example is not actually reproducible because you are not providing the data within the reprex itself, and you are not including library calls.

I think this is what you are trying to do

library(tidyverse, quietly = TRUE)
YearsPooled <- data.frame(stringsAsFactors=FALSE,
                          Year = as.factor(c(2015, 2015, 2015, 2015, 2015, 2015, 2015, 2015, 2015,
                                   2015, 2015, 2015, 2016, 2016, 2016, 2016, 2016, 2016,
                                   2016, 2016, 2016, 2016, 2016, 2016, 2016, 2016, 2016,
                                   2016, 2016, 2016, 2016, 2016, 2016, 2016, 2017, 2017,
                                   2017, 2017, 2017, 2017, 2017, 2017, 2017, 2017, 2015, 2015,
                                   2015, 2015, 2015, 2015, 2015, 2016, 2016, 2016, 2016,
                                   2016, 2016, 2016, 2016, 2016, 2016, 2016, 2016, 2016,
                                   2016, 2016, 2016, 2016, 2016, 2017, 2017, 2017)),
                          Lake = c("A6", "A6", "A6", "A6", "A6", "A8", "A8", "D7", "D7",
                                   "D7", "D7", "D7", "A6", "A6", "A6", "A8", "A8", "A8",
                                   "A8", "A8", "A8", "B2", "B2", "B2", "B2", "B2", "B2",
                                   "D7", "D7", "D7", "D7", "D7", "D7", "D7", "B2", "B2",
                                   "B2", "B2", "B2", "D7", "D7", "D7", "D7", "D7", "A6", "A8",
                                   "A8", "D7", "D7", "D7", "D7", "A6", "A6", "A6", "A6",
                                   "A6", "A8", "A8", "A8", "A8", "A8", "A8", "A8", "B2",
                                   "B2", "D7", "D7", "D7", "D7", "A6", "B2", "D7"),
                          LogWeight = log(c(0.0915, 0.166, 0.201, 0.1416, 0.1913, 0.365, 0.534,
                                           0.182, 0.1951, 0.2137, 0.2729, 0.3845, 0.0534, 0.7558,
                                           0.6106, 0.1219, 0.1249, 0.2519, 0.4617, 0.415, 0.9104,
                                           0.0643, 0.0807, 0.1435, 0.0292, 0.0689, 0.923, 0.706,
                                           0.761, 0.506, 0.579, 0.689, 0.562, 0.6049, 0.264, 0.3025,
                                           0.2666, 0.3032, 0.3573, 0.6741, 0.7112, 0.7173, 0.7104,
                                           0.9858, 0.5854, 0.5298, 0.8723, 0.5171, 0.55, 0.659,
                                           0.5147, 1.3292, 0.3911, 0.696, 0.9489, 0.6278, 0.1397,
                                           0.2709, 0.72, 0.6535, 0.8723, 0.9278, 1.1383, 0.13, 1.1902,
                                           0.2864, 0.4656, 0.3965, 1.1652, 1.7541, 0.8721, 0.5807)),
                          LogLength = log(c(29, 34, 36, 33, 35, 41, 48, 35, 34, 34, 39, 44, 24,
                                                48, 48, 29, 31, 37, 45, 42, 51, 24, 27, 30, 19, 23,
                                                52, 48, 54, 53, 57, 47, 55, 51, 35, 37, 36, 37.5, 41, 53,
                                                49, 51, 50, 59, 47, 48, 51, 46, 47, 46, 46, 61, 45, 47,
                                                55, 44, 34, 35, 51, 49, 53, 57, 57, 36, 57, 42, 41, 45,
                                                58, 68, 54, 55)),
                          K = c(0.375169134, 0.42234887, 0.430812757, 0.394022873,
                                0.446180758, 0.529591852, 0.482855903, 0.424489796,
                                0.496387136, 0.543710564, 0.460054957, 0.451375845,
                                0.386284722, 0.683412905, 0.552119502, 0.499815491,
                                0.419254137, 0.497305194, 0.506666667, 0.560144693, 0.686312203,
                                0.465133102, 0.409998476, 0.531481481, 0.425718035,
                                0.566285855, 0.656434911, 0.638382523, 0.483285068,
                                0.339877886, 0.312646806, 0.663629446, 0.337791134, 0.456008624,
                                0.61574344, 0.597200561, 0.571416324, 0.574957037,
                                0.51841964, 0.452789887, 0.604510026, 0.540742248, 0.56832,
                                0.479990651, 0.563844235, 0.47905816, 0.657590218,
                                0.531252568, 0.529747744, 0.677036246, 0.528786883,
                                0.585599676, 0.429190672, 0.67037169, 0.570338092, 0.736992863,
                                0.35543456, 0.631836735, 0.542777665, 0.555465835,
                                0.58591992, 0.500990858, 0.614656062, 0.278635117, 0.642680879,
                                0.386567325, 0.675556071, 0.435116598, 0.597195457,
                                0.557863067, 0.553840878, 0.349030804),
                          Parasites = as.factor(c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
                                                   0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
                                                   0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
                                                   1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1)) 
                          )
ggplot(YearsPooled, aes(x=LogLength, y=LogWeight, colour=Year, linetype=Parasites)) + 
    geom_point(size=2, aes(shape = Lake)) +
    stat_smooth(method = "lm", col = "black", size=0.25, se=FALSE, fullrange=TRUE) +
    theme_bw() +
    theme(plot.background = element_blank(), panel.grid.minor = element_blank(), panel.border = element_blank()) + 
    theme(axis.line = element_line(color = 'black')) +
    guides(shape=guide_legend(title=NULL), linetype=guide_legend(title=NULL)) +
    scale_color_grey("Year") +
    xlab("Length ln(mm)") +
    ylab("Weight ln(g)") +
    scale_shape_manual("Lake", values=c(15,16, 17, 18)) +
    NULL

^{Created on 2019-01-28 by the reprex package (v0.2.1)}

JVanslinger25 · January 28, 2019, 10:02pm

Hi, wow, that was exactly what I wanted to do, may I ask how your code worked and not mine, I did try changing a number of statements to allow it to code lake for shape. But what ended up happening is lake was treated for regressions so I got two regressions for each lake... Thanks!

andresrcs · January 28, 2019, 10:08pm

This is the line that makes the trick, you just have to map your variable to an aesthetic at the geom level.

JVanslinger25 · January 28, 2019, 10:09pm

Appreciate the help! thanks!

andresrcs · January 28, 2019, 10:09pm

If your question's been answered (even by you!), would you mind choosing a solution? It helps other people see which questions still need help, or find solutions if they have similar problems. Here’s how to do it:

system · February 4, 2019, 10:10pm

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.