library(readxl)
library(ggplot2)Load the necessary packages
Import the data from the Excel file
data <- read_excel("/Users/jal4510/Desktop/MRes/Applied Crop Sciences/SPAD_PPFD_FW.xlsx", sheet = "RED+BLUE")View the first few rows of the data to confirm it’s loaded correctly
head(data)# A tibble: 6 × 4
`Plant Tube` PPFD SPAD FW
<chr> <dbl> <dbl> <dbl>
1 A 141. 44 84
2 A 265. 33 24
3 A 202. 34.8 20
4 B 151. 28 92
5 B 319. 29.8 20
6 B 241. 22.8 46summary(data) Plant Tube PPFD SPAD FW
Length:12 Min. :138.4 Min. :22.4 Min. :13.00
Class :character 1st Qu.:151.1 1st Qu.:23.2 1st Qu.:20.00
Mode :character Median :221.8 Median :27.8 Median :26.00
Mean :232.0 Mean :28.5 Mean :43.25
3rd Qu.:278.4 3rd Qu.:30.6 3rd Qu.:60.75
Max. :413.6 Max. :44.0 Max. :99.00 Select columns X and Y for the correlation test
Perform Pearson correlation test
# Select columns for the correlation test
x <- data$PPFD
y <- data$SPAD
z <- data$FW
# Perform Pearson correlation tests for each pair of variables
cor_test_ppfd_spad <- cor.test(x, y) # Correlation between PPFD and SPAD
cor_test_ppfd_fw <- cor.test(x, z) # Correlation between PPFD and FW
cor_test_spad_fw <- cor.test(y, z) # Correlation between SPAD and FW
# Print the correlation test results
print(cor_test_ppfd_spad)
Pearson's product-moment correlation
data: x and y
t = -1.4371, df = 10, p-value = 0.1812
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
-0.7981248 0.2100443
sample estimates:
cor
-0.4137304 print(cor_test_ppfd_fw)
Pearson's product-moment correlation
data: x and z
t = -3.08, df = 10, p-value = 0.01164
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
-0.9080289 -0.2065196
sample estimates:
cor
-0.6977255 print(cor_test_spad_fw)
Pearson's product-moment correlation
data: y and z
t = 1.3013, df = 10, p-value = 0.2223
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
-0.247376 0.783367
sample estimates:
cor
0.3805524 Create a scatter plot with a linear regression line
# Create a scatter plot with a linear regression line, encoding FW with color
ggplot(data, aes(x = FW, y = PPFD, color = FW)) +
geom_point() + # Create scatter plot
geom_smooth(method = "lm", se = FALSE, color = "navy") + # Add linear regression line
labs(
title = paste("Scatter Plot of PPFD vs. FW\nCorrelation = ", round(cor(data$PPFD, data$FW), 2),
"\nP-value = ", round(cor_test_ppfd_spad$p.value, 4)),
x = "PPFD",
y = "FW"
) +
theme_minimal() +
scale_color_gradient(low = "blue", high = "red") # Color gradient for FW`geom_smooth()` using formula = 'y ~ x'
For One-Way Anova
# If you're interested in grouping 'PPFD' by different ranges of SPAD, consider discretizing SPAD.
# For example, here, I'll bin SPAD into categories for illustration
data$FW_group <- cut(data$FW, breaks = 3, labels = c("Low", "Medium", "High"))
print(data)# A tibble: 12 × 5
`Plant Tube` PPFD SPAD FW FW_group
<chr> <dbl> <dbl> <dbl> <fct>
1 A 141. 44 84 High
2 A 265. 33 24 Low
3 A 202. 34.8 20 Low
4 B 151. 28 92 High
5 B 319. 29.8 20 Low
6 B 241. 22.8 46 Medium
7 C 151. 27.6 53 Medium
8 C 319. 23.2 13 Low
9 C 241. 22.4 27 Low
10 D 200. 23.2 16 Low
11 D 414. 24 25 Low
12 D 138. 29.2 99 High # Perform One-Way ANOVA: PPFD as response and SPAD group as factor
anova_result <- aov(PPFD ~ FW_group, data = data)
# Step 4: View the ANOVA table to see if there's a significant difference
summary(anova_result) Df Sum Sq Mean Sq F value Pr(>F)
FW_group 2 42240 21120 4.85 0.0372 *
Residuals 9 39195 4355
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1# Step 5: (Optional) Perform a Tukey post-hoc test if ANOVA is significant
# Tukey's test is useful for pairwise comparison of groups
tukey_result <- TukeyHSD(anova_result)
summary(tukey_result) Length Class Mode
FW_group 12 -none- numeric# Step 6: Create a boxplot to visualize the group differences
ggplot(data, aes(x = FW_group, y = PPFD, fill = FW_group)) +
geom_boxplot() +
labs(title = "One-Way ANOVA: SPAD Group vs PPFD",
x = " FW Group",
y = "PPFD") +
theme_minimal()
FOR SUNLIGHT
# Load the necessary packages
library(readxl)
library(ggplot2)
# Import the data from the Excel file
data <- read_excel("/Users/jal4510/Desktop/MRes/Applied Crop Sciences/SPAD_PPFD_FW.xlsx", sheet = "Sunlight")
# View the first few rows of the data to confirm it's loaded correctly
head(data)# A tibble: 6 × 4
`Plant Tube` PPFD SPAD FW
<chr> <chr> <dbl> <dbl>
1 A 23.75 20 1
2 A 11.25 27 4
3 A 16.25 12.2 1
4 A 14.5 15.8 1
5 B 31.5 11.5 1
6 B 23.75 17.5 2summary(data) Plant Tube PPFD SPAD FW
Length:16 Length:16 Min. :10.00 Min. :0.500
Class :character Class :character 1st Qu.:15.06 1st Qu.:0.975
Mode :character Mode :character Median :17.55 Median :1.000
Mean :18.04 Mean :1.219
3rd Qu.:20.70 3rd Qu.:1.000
Max. :27.00 Max. :4.000 # Select columns for the correlation test
data$PPFD <- as.numeric(data$PPFD)Warning: NAs introduced by coercionx <- data$PPFD
y <- data$SPAD
z <- data$FW
# Perform Pearson correlation tests for each pair of variables
cor_test_ppfd_spad <- cor.test(x, y) # Correlation between PPFD and SPAD
cor_test_ppfd_fw <- cor.test(x, z) # Correlation between PPFD and FW
cor_test_spad_fw <- cor.test(y, z) # Correlation between SPAD and FW
# Print the correlation test results
print(cor_test_ppfd_spad)
Pearson's product-moment correlation
data: x and y
t = -0.37296, df = 13, p-value = 0.7152
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
-0.5843546 0.4321487
sample estimates:
cor
-0.1028916 print(cor_test_ppfd_fw)
Pearson's product-moment correlation
data: x and z
t = -2.3707, df = 13, p-value = 0.03389
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
-0.82849325 -0.05168065
sample estimates:
cor
-0.5493987 print(cor_test_spad_fw)
Pearson's product-moment correlation
data: y and z
t = 1.1058, df = 14, p-value = 0.2874
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
-0.2469797 0.6831509
sample estimates:
cor
0.2834251 # Create a scatter plot with a linear regression line, encoding FW with color
ggplot(data, aes(x = PPFD, y = FW, color = FW)) +
geom_point() + # Create scatter plot
geom_smooth(method = "lm", se = FALSE, color = "navy") + # Add linear regression line
labs(
title = paste("Scatter Plot of PPFD vs. FW\nCorrelation = ", round(cor(data$PPFD, data$FW), 2),
"\nP-value = ", round(cor_test_ppfd_fw$p.value, 4)), # Use correct p-value
x = "PPFD",
y = "FW"
) +
theme_minimal() +
scale_color_gradient(low = "blue", high = "red") # Color gradient for FW`geom_smooth()` using formula = 'y ~ x'Warning: Removed 1 row containing non-finite outside the scale range
(`stat_smooth()`).Warning: Removed 1 row containing missing values or values outside the scale range
(`geom_point()`).
# One-Way ANOVA: Group FW into categories
data$FW_group <- cut(data$FW, breaks = 3, labels = c("Low", "Medium", "High"))
print(data)# A tibble: 16 × 5
`Plant Tube` PPFD SPAD FW FW_group
<chr> <dbl> <dbl> <dbl> <fct>
1 A 23.8 20 1 Low
2 A 11.2 27 4 High
3 A 16.2 12.2 1 Low
4 A 14.5 15.8 1 Low
5 B 31.5 11.5 1 Low
6 B 23.8 17.5 2 Medium
7 B 26 10 2 Medium
8 B 25 16 1 Low
9 C 31.5 22.8 1 Low
10 C 23.8 26.6 1 Low
11 C 26 13 0.9 Low
12 C 25 17.6 0.6 Low
13 D 32 24.4 1 Low
14 D NA 19.6 0.5 Low
15 D 43.8 17 0.5 Low
16 D 35.4 17.6 1 Low # Perform One-Way ANOVA: PPFD as response and FW_group as factor
anova_result <- aov(PPFD ~ FW_group, data = data)
# View the ANOVA table to see if there's a significant difference
summary(anova_result) Df Sum Sq Mean Sq F value Pr(>F)
FW_group 2 242.3 121.17 2.023 0.175
Residuals 12 718.7 59.89
1 observation deleted due to missingness# Optional: Perform a Tukey post-hoc test if ANOVA is significant
tukey_result <- TukeyHSD(anova_result)
summary(tukey_result) Length Class Mode
FW_group 12 -none- numeric# Create a boxplot to visualize the group differences
ggplot(data, aes(x = FW_group, y = PPFD, fill = FW_group)) +
geom_boxplot() +
labs(title = "One-Way ANOVA: FW Group vs PPFD",
x = "FW Group",
y = "PPFD") +
theme_minimal()Warning: Removed 1 row containing non-finite outside the scale range
(`stat_boxplot()`).
For Nutrient
Load the necessary packages
library(readxl)
library(ggplot2)Import the data from the Excel file
data <- read_excel("/Users/jal4510/Desktop/MRes/Applied Crop Sciences/SPAD_PPFD_FW.xlsx", sheet = "Nutrient")View the first few rows of the data to confirm it’s loaded correctly
head(data)# A tibble: 6 × 4
Treatment PPFD SPAD FW
<dbl> <dbl> <dbl> <dbl>
1 1 43.2 19.1 1.2
2 1 43.2 17.2 1.2
3 2 41.7 21.5 2.2
4 2 41.7 24 2.2
5 3 45.6 19.6 2.4
6 3 45.6 20 2.4summary(data) Treatment PPFD SPAD FW
Min. :1.00 Min. :34.40 Min. :17.25 Min. :1.000
1st Qu.:2.75 1st Qu.:40.20 1st Qu.:20.82 1st Qu.:1.583
Median :4.50 Median :41.55 Median :23.40 Median :2.300
Mean :4.50 Mean :41.17 Mean :23.52 Mean :2.176
3rd Qu.:6.25 3rd Qu.:43.48 3rd Qu.:25.23 3rd Qu.:2.800
Max. :8.00 Max. :45.60 Max. :32.60 Max. :3.100 Select columns X and Y for the correlation test
Perform Pearson correlation test
# Select columns for the correlation test
data$PPFD <- as.numeric(data$PPFD)
x <- data$PPFD
y <- data$SPAD
z <- data$FW
# Perform Pearson correlation tests for each pair of variables
cor_test_ppfd_spad <- cor.test(x, y) # Correlation between PPFD and SPAD
cor_test_ppfd_fw <- cor.test(x, z) # Correlation between PPFD and FW
cor_test_spad_fw <- cor.test(y, z) # Correlation between SPAD and FW
# Print the correlation test results
print(cor_test_ppfd_spad)
Pearson's product-moment correlation
data: x and y
t = -4.1433, df = 14, p-value = 0.0009945
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
-0.9049448 -0.3899008
sample estimates:
cor
-0.7421638 print(cor_test_ppfd_fw)
Pearson's product-moment correlation
data: x and z
t = 1.8752, df = 14, p-value = 0.08178
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
-0.06126951 0.77223697
sample estimates:
cor
0.4480436 print(cor_test_spad_fw)
Pearson's product-moment correlation
data: y and z
t = -0.71749, df = 14, p-value = 0.4849
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
-0.6256264 0.3390293
sample estimates:
cor
-0.1883258 Create a scatter plot with a linear regression line
# Create a scatter plot with a linear regression line, encoding FW with color
ggplot(data, aes(x = PPFD, y = FW, color = FW)) +
geom_point() + # Create scatter plot
geom_smooth(method = "lm", se = FALSE, color = "blue") + # Add linear regression line
labs(
title = paste("Scatter Plot of PPFD vs. SPAD\nCorrelation = ", round(cor(x, y), 2),
"\nP-value = ", round(cor_test_ppfd_spad$p.value, 4)),
x = "PPFD",
y = "FW"
) +
theme_minimal() +
scale_color_gradient(low = "blue", high = "red") # Color gradient for FW`geom_smooth()` using formula = 'y ~ x'
For One-Way Anova
# If you're interested in grouping 'PPFD' by different ranges of SPAD, consider discretizing SPAD.
# For example, here, I'll bin SPAD into categories for illustration
data$FW_group <- cut(data$FW, breaks = 3, labels = c("Low", "Medium", "High"))
print(data)# A tibble: 16 × 5
Treatment PPFD SPAD FW FW_group
<dbl> <dbl> <dbl> <dbl> <fct>
1 1 43.2 19.1 1.2 Low
2 1 43.2 17.2 1.2 Low
3 2 41.7 21.5 2.2 Medium
4 2 41.7 24 2.2 Medium
5 3 45.6 19.6 2.4 Medium
6 3 45.6 20 2.4 Medium
7 4 44.3 21.1 3.1 High
8 4 44.3 23 3.1 High
9 5 41.4 23.3 3.1 High
10 5 41.4 23.5 3.1 High
11 6 41 23.6 2.7 High
12 6 41 27.5 2.7 High
13 7 37.8 25.6 1 Low
14 7 37.8 32.6 1 Low
15 8 34.4 25.1 1.71 Medium
16 8 34.4 29.6 1.71 Medium # Perform One-Way ANOVA: PPFD as response and SPAD group as factor
anova_result <- aov(PPFD ~ FW_group, data = data)
# Step 4: View the ANOVA table to see if there's a significant difference
summary(anova_result) Df Sum Sq Mean Sq F value Pr(>F)
FW_group 2 10.76 5.382 0.408 0.673
Residuals 13 171.43 13.187 # Step 5: (Optional) Perform a Tukey post-hoc test if ANOVA is significant
# Tukey's test is useful for pairwise comparison of groups
tukey_result <- TukeyHSD(anova_result)
summary(tukey_result) Length Class Mode
FW_group 12 -none- numeric# Step 6: Create a boxplot to visualize the group differences
ggplot(data, aes(x = FW_group, y = PPFD, fill = FW_group)) +
geom_boxplot() +
labs(title = "One-Way ANOVA: SPAD Group vs PPFD",
x = " FW Group",
y = "PPFD") +
theme_minimal()
For Red+Blue vs Sunlight
# Load the necessary packages
library(readxl)
library(ggplot2)
# Import the data from the Excel file
data <- read_excel("/Users/jal4510/Desktop/MRes/Applied Crop Sciences/SPAD_PPFD_FW.xlsx", sheet = "RED+BLUE vs Sunlight")
# View the first few rows of the data to confirm it's loaded correctly
head(data)# A tibble: 6 × 7
`Plant Tube` PPFD SPAD FW PPFD_Sunlight SPAD_Sunlight FW_Sunlight
<chr> <dbl> <dbl> <dbl> <chr> <dbl> <dbl>
1 A 141. 44 141. 23.75 20 1
2 A 265. 33 265. 11.25 27 4
3 A 202. 34.8 202. 16.25 12.2 1
4 B 151. 28 151. 31.5 11.5 1
5 B 319. 29.8 319. 23.75 17.5 2
6 B 241. 22.8 241. 26 10 2summary(data) Plant Tube PPFD SPAD FW
Length:12 Min. :138.4 Min. :22.4 Min. :138.4
Class :character 1st Qu.:151.1 1st Qu.:23.2 1st Qu.:151.1
Mode :character Median :221.8 Median :27.8 Median :221.8
Mean :232.0 Mean :28.5 Mean :232.0
3rd Qu.:278.4 3rd Qu.:30.6 3rd Qu.:278.4
Max. :413.6 Max. :44.0 Max. :413.6
PPFD_Sunlight SPAD_Sunlight FW_Sunlight
Length:12 Min. :10.00 Min. :0.500
Class :character 1st Qu.:12.81 1st Qu.:0.975
Mode :character Median :18.55 Median :1.000
Mean :18.47 Mean :1.325
3rd Qu.:23.20 3rd Qu.:1.250
Max. :27.00 Max. :4.000 # Select columns for the correlation test
x <- data$FW
y <- data$FW_Sunlight
# Perform Pearson correlation tests for each pair of variables
cor_test_FW_FW_Sunlight <- cor.test(x, y)
# Print the correlation test results
print(cor_test_FW_FW_Sunlight)
Pearson's product-moment correlation
data: x and y
t = 0.52635, df = 10, p-value = 0.6101
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
-0.4523372 0.6745295
sample estimates:
cor
0.1641868 # Create the FW_group column to categorize the 'FW' variable into 3 groups
data$FW_group <- cut(data$FW, breaks = 3, labels = c("Low", "Medium", "High"))
print(data)# A tibble: 12 × 8
`Plant Tube` PPFD SPAD FW PPFD_Sunlight SPAD_Sunlight FW_Sunlight
<chr> <dbl> <dbl> <dbl> <chr> <dbl> <dbl>
1 A 141. 44 141. 23.75 20 1
2 A 265. 33 265. 11.25 27 4
3 A 202. 34.8 202. 16.25 12.2 1
4 B 151. 28 151. 31.5 11.5 1
5 B 319. 29.8 319. 23.75 17.5 2
6 B 241. 22.8 241. 26 10 2
7 C 151. 27.6 151. 31.5 22.8 1
8 C 319. 23.2 319. 23.75 26.6 1
9 C 241. 22.4 241. 26 13 0.9
10 D 200. 23.2 200. 32 24.4 1
11 D 414. 24 414. 39,75 19.6 0.5
12 D 138. 29.2 138. 43.75 17 0.5
# ℹ 1 more variable: FW_group <fct># Create the FW_Sunlight_Group column (same as FW_group but using FW for sunlight exposure categorization)
data$FW_Sunlight_Group <- cut(data$FW, breaks = 3, labels = c("Low", "Medium", "High"))
print(data)# A tibble: 12 × 9
`Plant Tube` PPFD SPAD FW PPFD_Sunlight SPAD_Sunlight FW_Sunlight
<chr> <dbl> <dbl> <dbl> <chr> <dbl> <dbl>
1 A 141. 44 141. 23.75 20 1
2 A 265. 33 265. 11.25 27 4
3 A 202. 34.8 202. 16.25 12.2 1
4 B 151. 28 151. 31.5 11.5 1
5 B 319. 29.8 319. 23.75 17.5 2
6 B 241. 22.8 241. 26 10 2
7 C 151. 27.6 151. 31.5 22.8 1
8 C 319. 23.2 319. 23.75 26.6 1
9 C 241. 22.4 241. 26 13 0.9
10 D 200. 23.2 200. 32 24.4 1
11 D 414. 24 414. 39,75 19.6 0.5
12 D 138. 29.2 138. 43.75 17 0.5
# ℹ 2 more variables: FW_group <fct>, FW_Sunlight_Group <fct># Perform pairwise t-tests between the groups for 'PPFD' as the response variable
# Using 'FW_Sunlight_Group' as the factor for grouping
t_test_result_FW <- pairwise.t.test(data$PPFD, data$FW_Sunlight_Group)
# Print the pairwise t-test results with adjusted p-values
print(t_test_result_FW)
Pairwise comparisons using t tests with pooled SD
data: data$PPFD and data$FW_Sunlight_Group
Low Medium
Medium 0.00082 -
High 0.00026 0.00551
P value adjustment method: holm