Related
I have a dataframe (df1) with a column "PartcipantID". Some ParticipantIDs are wrong and should be replaced with the correct ParticipantID. I have another dataframe (df2) where all Participant IDs appear in columns Goal_ID to T4. The Participant IDs in column "Goal_ID" are the correct IDs.
Now I want to replace all ParticipantIDs in df1 with all Goal_ID ParticipantIDs from df2.
This is my original dataframe (df1):
structure(list(Partcipant_ID = c("AA_SH_RA_91", "AA_SH_RA_91",
"AB_BA_PR_93", "AB_BH_VI_90", "AB_BH_VI_90", "AB_SA_TA_91", "AJ_BO_RA_92",
"AJ_BO_RA_92", "AK_SH_HA_91", "AL_EN_RA_95", "AL_MA_RA_95", "AL_SH_BA_99",
"AM_BO_AB_49", "AM_BO_AB_94", "AM_BO_AB_94", "AM_BO_AB_94", "AN_JA_AN_91",
"AN_KL_GE_11", "AN_KL_WO_91", "AN_MA_DI_95", "AN_MA_DI_95", "AN_SE_RA_95",
"AN_SE_RA_95", "AN_SI_RA_97", "AN_SO_PU_94", "AN_SU_RA_91", "AR_BO_RA_92",
"AR_KA_VI_94", "AR_KA_VI_94", "AS_AR_SO_90", "AS_AR_SU_95", "AS_KU_SO_90",
"AS_MO_AS_97", "AW_SI_OJ_97", "AW_SI_OJ_97", "AY_CH_SU_97", "BH_BE_LD_84",
"BH_BE_LI_83", "BH_BE_LI_83", "BH_BE_LI_84", "BH_KO_SA_87", "BH_PE_AB_89",
"BH_YA_SA_87", "BI_CH_PR_94", "BI_CH_PR_94"), Start_T2 = structure(c(NA,
NA, NA, NA, 1579514871, 1576658745, NA, 1579098225, NA, NA, 1576663067,
1576844759, NA, 1577330639, NA, NA, 1576693930, NA, NA, NA, NA,
NA, NA, NA, NA, NA, NA, 1577718380, 1577718380, 1577454467, NA,
NA, 1576352237, NA, NA, NA, NA, 1576420656, 1576420656, NA, NA,
1578031772, 1576872938, NA, NA), class = c("POSIXct", "POSIXt"
), tzone = "UTC"), End_T2 = structure(c(NA, NA, NA, NA, 1579515709,
1576660469, NA, 1579098989, NA, NA, 1576693776, 1576845312, NA,
1577331721, NA, NA, 1576694799, NA, NA, NA, NA, NA, NA, NA, NA,
NA, NA, 1577719049, 1577719049, 1577455167, NA, NA, 1576352397,
NA, NA, NA, NA, 1576421607, 1576421607, NA, NA, 1578032408, 1576873875,
NA, NA), class = c("POSIXct", "POSIXt"), tzone = "UTC")), row.names = c(NA,
45L), class = "data.frame")
And this is the reference data frame (df2):
structure(list(Goal_ID = c("AJ_BO_RA_92", "AL_EN_RA_95", "AM_BO_AB_49",
"AS_KU_SO_90", "BH_BE_LI_84", "BH_YA_SA_87", "BI_CH_PR_94", "BI_CH_PR_94"
), T2 = c("AJ_BO_RA_92", "AL_MA_RA_95", "AM_BO_AB_94", "AS_AR_SO_90",
"BH_BE_LI_83", "BH_YA_SA_87", "BI_NA_PR_94", "BI_NA_PR_94"),
T3 = c("AR_BO_RA_92", "AL_MA_RA_95", "AM_BO_AB_94", NA, "BH_BE_LI_83",
NA, "BI_CH_PR_94", "BI_CH_PR_94"), T4 = c("AJ_BO_RA_92",
"AL_MA_RA_95", "AM_BO_AB_94", NA, "BH_BE_LI_83", "BH_KO_SA_87",
"BI_CH_PR_94", "BI_CH_PR_94")), row.names = c(NA, -8L), class = c("tbl_df",
"tbl", "data.frame"))
For example, in my df1, I want
"AR_BO_RA_92" to be replaced by "AJ_BO_RA_92";
"AL_MA_RA_95" to be replaced by "AL_EN_RA_95";
"AM_BO_AB_94" to be replaced by "AM_BO_AB_49"
and so on...
I thought about using string_replace and I started with this:
df1$Partcipant_ID <- str_replace(df1$Partcipant_ID, "AR_BO_RA_92", "AJ_BO_RA_92")
But that is of course very unefficient because I have so many replacements and it would be nice to make use of my reference data frame. I just cannot figure it out myself.
I hope this is understandable. Please ask if you need additional information.
Thank you so much already!
You can use match to find where the string is located and excange those which have been found and are not NA like:
i <- match(df1$Partcipant_ID, unlist(df2[-1])) %% nrow(df2)
j <- !is.na(i)
df1$Partcipant_ID[j] <- df2$Goal_ID[i[j]]
df1$Partcipant_ID
# [1] "AA_SH_RA_91" "AA_SH_RA_91" "AB_BA_PR_93" "AB_BH_VI_90" "AB_BH_VI_90"
# [6] "AB_SA_TA_91" "AJ_BO_RA_92" "AJ_BO_RA_92" "AK_SH_HA_91" "AL_EN_RA_95"
#[11] "AL_MA_RA_95" "AL_SH_BA_99" "AM_BO_AB_49" "AM_BO_AB_94" "AM_BO_AB_94"
#[16] "AM_BO_AB_94" "AN_JA_AN_91" "AN_KL_GE_11" "AN_KL_WO_91" "AN_MA_DI_95"
#[21] "AN_MA_DI_95" "AN_SE_RA_95" "AN_SE_RA_95" "AN_SI_RA_97" "AN_SO_PU_94"
#[26] "AN_SU_RA_91" "AR_BO_RA_92" "AR_KA_VI_94" "AR_KA_VI_94" "AS_AR_SO_90"
#[31] "AS_AR_SU_95" "AS_KU_SO_90" "AS_MO_AS_97" "AW_SI_OJ_97" "AW_SI_OJ_97"
#[36] "AY_CH_SU_97" "BH_BE_LD_84" "BH_BE_LI_83" "BH_BE_LI_83" "BH_BE_LI_84"
#[41] "BH_KO_SA_87" "BH_PE_AB_89" "BH_YA_SA_87" "BI_CH_PR_94" "BI_CH_PR_94"
I think this might work. Create a true look up table with a column of correct and incorrect codes. I.e. stack the columns, then join the subsequent df3 to df1 and use coalesce to create a new part_id. You spelt participant wrong, which made me feel more human I always do that.
library(dplyr)
df3 <- df2[1:2] %>%
bind_rows(df2[c(1,3)] %>% rename(T2 = T3),
df2[c(1,4)] %>% rename(T2 = T4)) %>%
distinct()
df1 %>%
left_join(df3, by = c("Partcipant_ID" = "T2")) %>%
mutate(Goal_ID = coalesce(Goal_ID, Partcipant_ID)) %>%
select(Goal_ID, Partcipant_ID, Start_T2, End_T2)
I am looking at correlations between many variables in my data stratified by gender. I was able to create a heatmap using code I found on StackOverflow, but I'm not sure how to add stars for significance to the cells. I would also like to cut the matrix in half to avoid redundancy.
Here's the code:
# Variables to correlate
anthro <- c("Visit_age", "HeightCm", "WeightKg", "BMI",
"NeckLengthCm", "NeckCircCm", "HeadCircCm", "NeckVolumeCm")
peak <- c("ExtensorPeak_Newtons", "FlexorPeak_Newtons",
"RightPeak_Newtons", "LeftPeak_Newtons")
avg <- c("ExtensorAVG_Newtons", "FlexorAVG_Newtons",
"RightAVG_Newtons", "LeftAVG_Newtons")
# Function for creation of multiple heatmaps using
# male/female and peak/avg neck strength
heatmap <- function(gender, strength){
# Create three new variables: var1, var2, corr
# where corr is correlation between the var1 and var2
corrs <- filter(data, Gender == gender) %>%
select(anthro, strength) %>%
as.matrix() %>%
cor(use = "pairwise.complete.obs") %>%
as.data.frame() %>%
rownames_to_column(var = "var1") %>%
gather("var2", "corr", -var1)
# Plot heatmap
ggplot(corrs, aes(var1, var2)) +
geom_tile(aes(fill = corr), color = "white") +
scale_fill_gradient(low = "white", high = "steelblue") +
geom_text(aes(label = round(corr, 1))) +
ggtitle(gender) +
labs(x = "", y = "") +
theme(plot.title = element_text(hjust = 0.5),axis.text.x =
element_text(angle = 30, hjust = 1))
}
# Create heatmaps
heatmap("Male", peak)
heatmap("Female", peak)
heatmap("Male", avg)
heatmap("Female", avg)
dput(head(data, 20)):
data <- structure(list(Gender = structure(c(2L, 2L, 2L, 1L, 1L, 1L, 1L,
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L), .Label = c("Male",
"Female"), class = "factor"), Visit_age = c(37, 38, 39, 22, 23,
24, 24, 20, 21, 21, 22, 22, 36, 37, 38, 38, 22, 42, 42, 43),
HeightCm = c(170, 170, 170, 182, 182, 182, 182, 177.8, 177.8,
177.8, 177.8, 177.8, 168, 168, 168, 168, 162.56, 164, 164,
164), WeightKg = c(63.18181, 58.63636, 60.45454, 70.90909,
77.72727, 75.45454, 80.45454, 78.86363, 81.36363, 80, 83.18181,
82.72727, 68.18181, 69.0909, 68.18181, 65, 69.0909, 48.18181,
50.45454, 47.72727), BMI = c(21.86222, 20.28939, 20.91852,
21.40716, 23.46554, 22.77941, 24.28889, 24.94671, 25.73752,
25.30617, 26.31266, 26.16888, 24.15739, 24.47948, 24.15739,
23.03004, 26.14529, 17.91412, 18.75912, 17.74511), NeckLengthCm = c(16,
16, 16, 14, 14, 14, 14, 16, 16, 16, 16, 16, 16, 16, 16, 16,
15, 15, 15, 15), NeckCircCm = c(35, 30, 32, 35, 34, 34, 36,
38, 39, 38, 40, 41, 39, 24, 36, 38, 34, 30, 29, 30), HeadCircCm = c(58,
58, 58, 56, 56, 56, 56, 57, 57, 57, 57, 57, 58, 58, 58, 58,
55, 52, 52, 52), NeckVolumeCm = c(1559.718, 1145.915, 1303.797,
1364.753, 1287.881, 1287.881, 1443.853, 1838.557, 1936.597,
1838.557, 2037.183, 2140.315, 1936.597, 733.3859, 1650.118,
1838.557, 1379.873, 1074.295, 1003.869, 1074.295), ExtensorPeak_Newtons = c(NA,
183.34, 145.96, NA, NA, 187.79, 153.525, NA, NA, 252.76,
227.395, 192.685, NA, NA, 168.21, 230.51, NA, NA, NA, 101.015
), FlexorPeak_Newtons = c(NA, 70.755, 68.975, NA, NA, 99.68,
112.585, NA, NA, 151.3, 136.615, 145.96, NA, NA, 97.9, 105.02,
NA, NA, NA, 53.4), RightPeak_Newtons = c(NA, 93.005, 125.935,
NA, NA, 85.885, 92.56, NA, NA, 102.35, 108.135, 108.135,
NA, NA, 74.315, 97.01, NA, NA, NA, 49.395), LeftPeak_Newtons = c(NA,
125.49, 131.275, NA, NA, 89.89, 99.68, NA, NA, 113.92, 121.93,
143.29, NA, NA, 59.185, 92.56, NA, NA, NA, 50.73), ExtensorAVG_Newtons = c(NA,
179.186637, 142.5483185, NA, NA, 178.445, 136.911637, NA,
NA, 242.97, 204.106637, 167.765, NA, NA, 161.09, 214.49,
NA, NA, NA, 95.081637), FlexorAVG_Newtons = c(NA, 68.2333185,
66.75, NA, NA, 87.516637, 100.125, NA, NA, 135.131637, 128.7533185,
138.84, NA, NA, 88.406637, 95.971637, NA, NA, NA, 51.62),
RightAVG_Newtons = c(NA, 85.1433185, 120.2983185, NA, NA,
75.65, 86.4783185, NA, NA, 96.7133185, 100.866637, 106.9483185,
NA, NA, 67.046637, 88.851637, NA, NA, NA, 47.7633185), LeftAVG_Newtons = c(NA,
121.93, 120.2983185, NA, NA, 74.315, 92.56, NA, NA, 110.656637,
111.546637, 130.83, NA, NA, 54.29, 88.11, NA, NA, NA, 48.801637
)), row.names = c(NA, -20L), class = c("tbl_df", "tbl", "data.frame"
))
I found an alternative way to resolve your problem on http://www.sthda.com/english/wiki/visualize-correlation-matrix-using-correlogram
Try to make a correlogram
library(corrplot)
# Correlation for Male
data_male <- data[data$Gender == "Male",]
M <- cor(data_male[,-1], use = "pairwise.complete.obs")
M <- round(M, 1)
#Significant correlation
p.mat <- cor(data_male[,-1])
# Plot the correlogram
col <- colorRampPalette(c("#BB4444", "#EE9988", "#FFFFFF", "#77AADD", "#4477AA"))
corrplot(M,
method="color",
col=col(200),
type="upper",
order="hclust",
addCoef.col = "black",
tl.col="black",
number.cex = 0.7,
tl.cex = 0.6,
tl.srt=45,
p.mat =p.mat,
sig.level = 0.5,
insig = "label_sig")
You can do the same thing for Female
data_female <- data[data$Gender == "Female",]
F <- cor(data_female[,-1], use = "pairwise.complete.obs")
F <- round(F, 1)
corrplot(F,
method="color",
col=col(200),
type="upper",
order="hclust",
addCoef.col = "black",
tl.col="black",
number.cex = 0.7,
tl.cex = 0.6,
tl.srt=45,
p.mat =p.mat,
sig.level = 0.5,
insig = "label_sig")
Instead of your current argument to geom_text(aes(label= ...)) use:
label = paste(round(corr,1), c(" ","*")[(abs(corr) <= .05)+1])
This will add a "*" when the absolute value of corr is below 0.05.
Look at the code of ggcorrplot::ggcorrplot to see how they handle filling only half a square tile plot.
I tried to generate a "forest plot" without summary estimates using the rmeta package. However, using ?forestplot and then starting from the description or the example does not help, I am always getting the same error. I would assume that it is a simple one that has to do with the matrix/vector lengths somewhat not lining up but I kept changing and adjusting and still cannot find the error...
Here is the example code:
tabletext<-cbind(c(NA, NA, NA, NA, NA, NA),
c(NA, NA, NA, NA, NA, NA),
c("variable1","subgroup","2nd", "3rd", "4th", "5th"),
c(NA,"mean","1.8683639", "2.5717301", "4.4966049, 9.0008054")
)
tabletext
png("forestplot.png")
forestplot(tabletext, mean = c(NA, NA, 1.8683639, 2.5717301, 4.4966049, 9.0008054), lower = c(NA, NA, 1.4604643, 2.0163468, 3.5197956, 6.9469213), upper = c(NA, NA, 2.3955105, 3.2897459, 5.7672966, 11.7288609),
is.summary = c(rep(FALSE, 6)), zero = 1, xlog=FALSE, boxsize=0.75, xticks = NULL, clip = c(0.9, 12))
dev.off()
Error message:
clip = c(0.9, 12))
Error in unit(rep(1, sum(widthcolumn)), "grobwidth", labels[[1]][widthcolumn]) :
'x' and 'units' must have length > 0
dev.off()
Any help is very much appreciated!
This works with the forestplot-package although you need to remove the xticks=NULL:
tabletext<-cbind(c(NA, NA, NA, NA, NA, NA),
c(NA, NA, NA, NA, NA, NA),
c("variable1","subgroup","2nd", "3rd", "4th", "5th"),
c(NA,"mean","1.8683639", "2.5717301", "4.4966049, 9.0008054")
)
png("forestplot.png")
forestplot(tabletext,
mean = c(NA, NA, 1.8683639, 2.5717301, 4.4966049, 9.0008054),
lower = c(NA, NA, 1.4604643, 2.0163468, 3.5197956, 6.9469213),
upper = c(NA, NA, 2.3955105, 3.2897459, 5.7672966, 11.7288609),
is.summary = c(rep(FALSE, 6)), zero = 1,
xlog=FALSE, boxsize=0.75, clip = c(0.9, 12))
dev.off()
Gives (I recommend some polishing before submitting for publishing):
I'm having trouble to create a proper boxplot of my dataset. All of the solutions on this platform don't work because their dataset all look different with variables against each other.
So I want to ask: how do I need to format my dataset if it only contains 3 variables and their measured values in 3 columns. In the boxplot examples here, they plot a variable against another one but here this is not the case right?
Using boxplot(data) gives me 3 boxplots. But I want to show the MEAN and also the population size on each boxplot. I don't know how to use the solution as they are all about ggplot2 or boxplot with variables against each other.
I know that this must be simple, but I think I'm plotting the boxplots on a bad method and that's why the solutions on this site don't work?
Data:
structure(list(Rest = c(3.479386607, 3.478445796, 2.52227462,
1.726115552, 3.917693859, 2.300840122), Peat = c(16.79515746,
22.76673699, 24.43289941, 15.64168939, 31.60459098, 16.2369787
), Top.culture = c(8.288, 8.732, 5.199, 6.539, 3.248, 10.156)), .Names = c("Rest",
"Peat", "Top.culture"), row.names = c(NA, 6L), class = "data.frame")
If text annotation is what is meant by 'show the mean and also the population size' then:
boxplot(dat)
text(1:3, 12.5, paste( "Mean= ",round(sapply(dat,mean, na.rm=TRUE), 2),
"\n N= ",
sapply(dat, function(x) length( x[!is.na(x)] ) )
) )
This used your more complex data-object from the other (duplicated) question.
dat <- structure(list(Rest = c(3.479386607, 3.478445796, 2.52227462, 1.726115552, 3.917693859, 2.300840122, 2.326307503, 2.344828287, 4.654278623, 3.68669447, 3.343706863, 0.712228306, 2.735897248, 1.936723375, 2.724260325, 2.069633651, 1.741484154, 2.304391217, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA), Peat = c(16.79515746, 22.76673699, 24.43289941, 15.64168939, 31.60459098, 16.2369787, 32.63285246, 35.91852324, 19.27802839, 21.78974576, 30.39119451, 35.4846573, 42.21807817, 42.00913743, 40.96996704, 19.85075354, 17.247096, 22.81689524, 43.35990368, 37.57273508, 23.76889902, 38.34604591, 20.98376674, 16.44173119, 17.27639888, NA, NA, NA, NA, NA, NA), Top.culture = c(8.288, 8.732, 5.199, 6.539, 3.248, 10.156, 3.436, 5.584, 4.483, 2.087, 3.28, 2.71, 2.196, 4.971, 4.475, 6.361, 5.49, 9.085, 3.52, 5.772, 9.308, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA)), .Names = c("Rest", "Peat", "Top.culture" ), class = "data.frame", row.names = c(NA, -31L))
I have a for-loop that initializes 3 vectors (launch_2012, amount, and one_week_bf) and creates a data frame. Then, it predicts a single week's of data and inserts it into vectors (amount and one_week_bf), and recreates the data.frame again; this process is looped 8 times. However, I can't seem to get the data.frame to update the new amounts. Would anyone be able to assist please?
for (i in 1:8) {
launch_2012 <- c(rep('bf', 5), 'launch', rep('af', 7))
amount <- c(7946, 6641, 5975, 5378, 5217, NA, NA, NA, NA, NA, NA, NA, NA)
one_week_bf <- c(NA, 7946, 6641, 5975, 5378, 5217, NA, NA, NA, NA, NA, NA, NA)
newdata <- data.frame(amount = amount, one_week_bf = one_week_bf, launch = launch_2012, week = week)
predicted <- predict(model0a, newdata)
amount[i+5] <- predicted[i+5]
one_week_bf[i+6] <- predicted[i+5]
View(newdata)
}
It's difficult to be sure since your example is not reproducible, but note that predict.lm(...) by default has na.action=na.pass, which means that any rows in newdata that have any NA values by default generate NA for the prediction. Since your first pass of newdata has NA in rows 6-13, predicted will have NA in those same elements. This means that amounts and one_week_bf will have NA in those elements, which in turn will generate the same newdata each time.
None of this should be in a for loop.
x <- data.frame("launch_2012" = c(rep('bf', 5), 'launch', rep('af', 7)),
"amount"=c(7946, 6641, 5975, 5378, 5217, NA, NA, NA, NA, NA, NA, NA, NA),
"one_week_bf"=c(NA, 7946, 6641, 5975, 5378, 5217, NA, NA, NA, NA, NA, NA, NA))
x$new_amount <- #the replacement from your predict vector
x$new_one_week_bf <- #the replacement from your predict vector
Note I have no idea what model0a does, so just gave what the new columns should be as whatever the resulting vector is from your predict function. This will add the new data as new columns