Last active
July 24, 2020 02:42
-
-
Save timedreamer/9a662cfe28a259d02d99426e165a4374 to your computer and use it in GitHub Desktop.
Convert network precision score to the edge weight.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
I used the precision curve only to decide the cutoff value. I chose a precision cutoff 0.2, then the correspondent normalized ranking is *~0.01*. The meaning of the normalized ranking is explained [here](https://github.com/takayasaito/precrec/issues/12). | |
Therefore, to calculate the **weight** value for precision 0.2, we got the following calculation: (rank-1)/(n-1) = 0.01. In this case, `n=nrow(dfg_ortho_label)`, so the rank is *649*. We went back to the *649* row of the `dfg_ortho_label`, the `weight` is 1.57. Therefore, we kept the edge that has weight higher than 1.57. | |
```{r, fig.width=5, fig.height=5, fig.align="center"} | |
scurve_os_b <- evalmod(scores = dfg_ortho_label$weight, labels = dfg_ortho_label$label, | |
mode = "basic") | |
sos_df_b <- fortify(scurve_os_b) | |
p2 <- ggplot(subset(sos_df_b, curvetype == "precision"), aes(x = x, y = y))+ | |
geom_point(color = "blue", size = 0.4)+ ylim(0:1) | |
p2 + geom_hline(yintercept = 0.2, color = "black", linetype = "dashed") + | |
geom_vline(xintercept = 0.022, color = "black", linetype = "dashed") + | |
xlab("normalized rank") + | |
ylab("Precision") + | |
geom_vline(xintercept = 0.015, color = "black", linetype = "dashed") | |
``` |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment