Question

I started to use ggplot in order to generate some graphs. I am using the very basic of the sample they have in their site for an “step graph”: http://had.co.nz/ggplot2/geom_step.html

I need something like the last example in that page. The issue I found is the example is in one DF and every Y has a different value in X. I have 3 sets to plot, all of them has the same X and Y values in all X. I don’t get how to produce the graphic in R. In excel I create the first line and add the other 2 as new data series but I am lost in ggplot. Any advise is really appreciated.

My plot commands are now as this (each of them with different df):

qplot(seq_along(x), x, data = df, geom="step", colour = trt)

Is there a way to have one df with all the values? Any example will be great. Thanks in advance for your time

EDIT>>>>

Here is some data to see:

head(df_nourbanas)
     x        trt
1    3 no urbanas
2   85 no urbanas
3  450 no urbanas
4  942 no urbanas
5 1362 no urbanas
6 1699 no urbanas

head(df_urbanas)
     x     trt
1   65 urbanas
2 1019 urbanas
3 2192 urbanas
4 3164 urbanas
5 3847 urbanas
6 4338 urbanas

head(df_todas)
     x   trt
1    3 todas
2  152 todas
3 1481 todas
4 3158 todas
5 4564 todas
6 5593 todas

And this is what I need to have (this is in excel) directly form R using the steps mode

Answer 1

You are having problems because you are using implicit information in the data set: the order the points are, which becomes the x-axis position. Make this explicit and things become easier.

df_nourbanas$idx <- seq(length=nrow(df_nourbanas))
df_urbanas$idx <- seq(length=nrow(df_urbanas))
df_todas$idx <- seq(length=nrow(df_todas))

Now these 3 data sets can be combined.

df_all <- rbind(df_nourbanas, df_urbanas, df_todas)

and plotting them is just

ggplot(df_all, aes(x=idx, y=x, colour=trt)) +
  geom_step()