Attaching package: 'lubridate'
The following objects are masked from 'package:base':
date, intersect, setdiff, union
library(palmerpenguins)library(patchwork)
Essentials:
1.) Read in https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2021/2021-05-04/water.csv check the format of the date column and change it using lubridate so it is correct
#read in the water datawater<-read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2021/2021-05-04/water.csv')
Rows: 473293 Columns: 13
── Column specification ────────────────────────────────────────────────────────
Delimiter: ","
chr (9): report_date, status_id, water_source, water_tech, facility_type, co...
dbl (4): row_id, lat_deg, lon_deg, install_year
ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.
water_1 <- waterhead(water_1)
# A tibble: 6 × 13
row_id lat_deg lon_deg repor…¹ statu…² water…³ water…⁴ facil…⁵ count…⁶ insta…⁷
<dbl> <dbl> <dbl> <chr> <chr> <chr> <chr> <chr> <chr> <dbl>
1 3957 8.07 38.6 04/06/… y <NA> <NA> <NA> Ethiop… NA
2 33512 7.37 40.5 08/04/… y Protec… <NA> Improv… Ethiop… 2019
3 35125 0.773 34.9 03/18/… y Protec… <NA> Improv… Kenya NA
4 37760 0.781 35.0 03/18/… y Boreho… <NA> Improv… Kenya NA
5 38118 0.779 35.0 03/18/… y Protec… <NA> Improv… Kenya NA
6 38501 0.308 34.1 03/19/… y Boreho… <NA> Improv… Kenya NA
# … with 3 more variables: installer <chr>, pay <chr>, status <chr>, and
# abbreviated variable names ¹report_date, ²status_id, ³water_source,
# ⁴water_tech, ⁵facility_type, ⁶country_name, ⁷install_year
water_1$report_date<-mdy(water_1$report_date) #converts our date column into a date/time object based on the format (order) of our date and time str(water_1)
spc_tbl_ [473,293 × 13] (S3: spec_tbl_df/tbl_df/tbl/data.frame)
$ row_id : num [1:473293] 3957 33512 35125 37760 38118 ...
$ lat_deg : num [1:473293] 8.073 7.374 0.773 0.781 0.779 ...
$ lon_deg : num [1:473293] 38.6 40.5 34.9 35 35 ...
$ report_date : Date[1:473293], format: "2017-04-06" "2020-08-04" ...
$ status_id : chr [1:473293] "y" "y" "y" "y" ...
$ water_source : chr [1:473293] NA "Protected Spring" "Protected Shallow Well" "Borehole" ...
$ water_tech : chr [1:473293] NA NA NA NA ...
$ facility_type: chr [1:473293] NA "Improved" "Improved" "Improved" ...
$ country_name : chr [1:473293] "Ethiopia" "Ethiopia" "Kenya" "Kenya" ...
$ install_year : num [1:473293] NA 2019 NA NA NA ...
$ installer : chr [1:473293] "Private-CRS" "WaterAid" NA NA ...
$ pay : chr [1:473293] NA NA NA NA ...
$ status : chr [1:473293] NA NA NA NA ...
- attr(*, "spec")=
.. cols(
.. row_id = col_double(),
.. lat_deg = col_double(),
.. lon_deg = col_double(),
.. report_date = col_character(),
.. status_id = col_character(),
.. water_source = col_character(),
.. water_tech = col_character(),
.. facility_type = col_character(),
.. country_name = col_character(),
.. install_year = col_double(),
.. installer = col_character(),
.. pay = col_character(),
.. status = col_character()
.. )
- attr(*, "problems")=<externalptr>
# date which was initially a character is now date object: it is read as date and thus easy to work with.
'data.frame': 77 obs. of 6 variables:
$ site : int 1 1 1 1 1 2 2 2 2 2 ...
$ type : chr "Back Reef" "Back Reef" "Back Reef" "Back Reef" ...
$ lat : int 3 3 3 3 3 3 3 3 3 3 ...
$ transect : int 1 2 3 4 5 1 2 3 4 5 ...
$ diver : int 4 4 4 5 5 4 4 4 5 5 ...
$ cc_percent: num 5.844 0.951 5.242 5.004 5.895 ...
3 Look at the data and generate a hypothesis to test (X, Y, and/or Z has no effect on coral cover (cc_percent), for example). cc_percent is the only numerical variable we care about here! Everything else is categorical. Write your hypothesis in BOLD below.
Hypothesis: the type of reef has no effect on the percent distribution of coral cover
\(H_0:\) the type of reef has no effect on the percent coral cover distribution
\(H_A:\) the type of reef has an effect on the percent coral cover distribution
4 Filter our the columns that you are not using for your hypothesis test
coral_fil<- coral_1[,c(2,6)]head(coral_fil)
type cc_percent
1 Back Reef 5.8435758
2 Back Reef 0.9505263
3 Back Reef 5.2423389
4 Back Reef 5.0040475
5 Back Reef 5.8954916
6 Patch Reef 5.2826190
5 Using the pipe, %>%, group your data and calcualte the mean(s) you need for your visual hypothesis test
coral_fil2 <- coral_fil %>%group_by(type) %>%#group = type of reef summarize(mean_cover =mean(cc_percent), sd_cover =sd(cc_percent), n=n(), se = sd_cover/sqrt(n)) head(coral_fil2)
# A tibble: 3 × 5
type mean_cover sd_cover n se
<chr> <dbl> <dbl> <int> <dbl>
1 Back Reef 11.9 9.52 29 1.77
2 Nearshore 3.16 2.25 18 0.531
3 Patch Reef 12.7 11.5 30 2.11
6 Graph your results! Means + errorbars required :) Make a nice, easy to see graph with clear labels and text
p <-ggplot(data = coral_fil2,aes(y = mean_cover, x = type, color ="type")) +geom_point()+labs(title ="% distribution of coral cover per reef", x="type", y ="cc_percent")+theme_bw()+theme(axis.text.x=element_text(angle =0, size =15, vjust =1), axis.title.y =element_text(size =16),axis.title.x =element_text(size =16))p+geom_errorbar(aes(ymin=mean_cover-se, ymax=mean_cover+se), width=.2,position=position_dodge(.9))
7 Assess your hypothesis! What does your graph show (Note: We did not do stats, so please do not say ‘significant’)
The percent distribution of coral cover is very different especially at Near shore (lower) as compared to Patch Reef and Back Reef. Even though the Patch Reef had seemingly the highest % coral cover, it also contains the highest error bars ( which shows that there is very high variation/spread of the data from the mean). Therefore, the type of Reef has an effect on the cc_percent distribution.
Depth
1: Read in intertidal transect data. View it, identify the columns that contain species/cover items and pivot from wide to long format!
2 Filter data such that we retain only ‘species’ that are animals. These include: Carinbus, Cancer Crabs, nucella, litt_obt, litt_lit, litt_sax. Everything else is either rock or algae. Please do this filter step AFTER you pivot to long format. Note: It is easier to do this when you are in wide format, but this is good practice! Ask questions if you need help :). Hint: The ‘|’ key is how you say ‘or’ in code :)
long_op2<-filter(long_op2, species =="carcinus"|species =="Cancer.crabs"|species =="nucella"|species =="litt_obt"|species =="litt_lit"|species =="litt_sax"|species =="semibal") head(long_op2)
# vertical line equals to OR#tidyverse version: df <- long_op %>%
2 Using the same data–> rename the factors in the trans_description column based on wave exposure. a_cobble_protected is low, c_flat_profile and d_semi_expos are both moderate, e_exposed is high. Hint: use ifelse(). Justin can help!
# makes a new column(input the dataframe$name of column)
3 Using the same data, make a simplified tidal height column. We want tidal height cat <4 to be ‘low’, >7 to be ‘high’, and in between to be ‘intermediate’. Hint: You can use if_else for this as well!
4 Using the same dataframe, build 2 new dataframes, one calculating mean and error (standard error) for semibal (barnacle) abundance and one doing the same for nucella (whelk) abundance by tidal height and wave exposure (trans_description)
5 Plot barnacle and snail abundance + error across tidal heights & wave exposures. You can plot them both on the same graphs (merging your dataframes could help) or you can make multiple graphs and patchwork them together.
p1 <-ggplot(data =long_op4,aes(y = mean_nucella, x = wave_exposure)) +geom_point()+labs(title ="Abundance of snail by wave exposure", y="mean_nucella abundance", x ="wave exposure")+theme_bw()+theme(axis.text.x=element_text(angle =0, size =15, vjust =1), axis.title.y =element_text(size =16),axis.title.x =element_text(size =16))p1+geom_errorbar(aes(ymin=mean_nucella, ymax=se_nucella), width=.2,position=position_dodge(.9)) +geom_jitter(data =long_op5,aes(y = counts, x = wave_exposure))
p2 <-ggplot(data =long_op3,aes(y = mean_semibal, x = wave_exposure)) +geom_point()+labs(title ="Abundance of whelk by wave exposure", y="mean_semibal abundance", x ="wave exposure")+theme_bw()+theme(axis.text.x=element_text(angle =0, size =15, vjust =1), axis.title.y =element_text(size =16),axis.title.x =element_text(size =16))p2+geom_errorbar(aes(ymin=mean_semibal, ymax=se_semibal), width=.2,position=position_dodge(.9)) +geom_jitter(data =long_op6,aes(y = counts, x = wave_exposure))
The mean abundance of snail is uniform and relatively similar at low and moderate wave exposure. However there is high variation at moderate wave exposure than at low wave exposure.The mean abundance of whelk is higher at lower wave exposure. There is high spread around the mean in both cases.
For both the snail and the whelk, there is zero mean abundance at high wave exposure.
Overall, the mean abundance of snails is generally higher than whelk distribution for (low and moderate) wave exposures, however, there is also very high spread of the data from the mean (high variation) for the whelk population
p3 <-ggplot(data =long_op4,aes(y = mean_nucella, x = tidal_ht_cat_num)) +geom_point()+labs(title ="Abundance of snail by tidal height", y="mean_nucella abundance", x ="tidal_ht_cat_num")+theme_bw()+theme(axis.text.x=element_text(angle =0, size =15, vjust =1), axis.title.y =element_text(size =16),axis.title.x =element_text(size =16))p3+geom_errorbar(aes(ymin=mean_nucella, ymax=se_nucella), width=.2,position=position_dodge(.9)) +geom_jitter(data =long_op5,aes(y = counts, x = tidal_ht_cat_num))
p4 <-ggplot(data =long_op3,aes(y = mean_semibal, x = tidal_ht_cat_num)) +geom_point()+labs(title ="Abundance of whelk by tidal height", y="mean_semibal abundance", x ="tidal_ht_cat_num")+theme_bw()+theme(axis.text.x=element_text(angle =0, size =15, vjust =1), axis.title.y =element_text(size =16),axis.title.x =element_text(size =16))p4+geom_errorbar(aes(ymin=mean_semibal, ymax=se_semibal), width=.2,position=position_dodge(.9)) +geom_jitter(data =long_op6,aes(y = counts, x = tidal_ht_cat_num))
There is higher distribution of snails at high wave exposure and no whelk at high wave exposure. on the other hand, there is high mean abundance of whelk at moderate tidal height than snail abundance. In addition, there is high spread around the mean for both the snail and whelk population at moderate tidal height.
6 Write a short interpretative statement. We didn’t run any stats, so avoid the word ‘significant.’ How do snail and whelk abundances appear to vary by tidal height and wave exposure? I do not need you to tell me anything about the ecology of the system, but feel free to do so if you’d like :) I just need you to interpret your graphs.
see explanation under the graphs above
I tried to patch my graphs, but when i did, the error bars and some data points were not visible anymore - it seemed like it the data became so condensed and i also failed at merging the data frames to one - so was not really able have them side by side as recommended.
