I don't use audacity, but my typical procedure for editing acoustic shows is:
First, apply a limiter/compression to the sections with loud clapping and crowd. If it is between-song clapping you can be very aggressive with setting the threshold for where the conpression kicks in. The clapping will sound a little weird if you are really compressing the heck out of it but I don't really care as it is not music. You could play around with the attack and release settings but I personally haven't doodled with them. If the clapping occurs during music you will need to do some more judicious application of your compressor so as to not overly affect the music, for these parts I use a soft limiter and set it to just compress the claps. Give that section a listen to make sure it sounds ok to you. When finished with compressing all the portions of crowd noise, my rule of thumb is that the loudest peaks in the file should be the music.
Then, I peak normalize the whole file to bring everything up in level.
This has worked pretty well for me when editing acoustic music, and made the recordings much more listenable. I am sure some others have different methods. Good luck!