Hi there.
I am not sure whether it is normal and usual to spend an entire day on elaborating the plot of given data set every day in a week.
Before, I was doing many sets of perturbative calculations using Mathematica, each taking about 3 hours. It took about 4 plus 2 weeks to get the correct expression.
After that, I have been installing and struggling with an unhelpful manual for a week.
Now I got the data, but I should learn how to use another tool that I am completely new to plot multiple graphs.

The point is, I couldn't find these things relevant/helpful in obtaining bits of knowledge in my field. It may make me a more skilled person but I doubt spending the whole days and weeks on those things would help me become a better scholar.

Looking at other people in the office involved in a meaningful calculation and reading multiple papers a day, I certainly could draw a big gap between theirs and mine. I asked one of my friends for an advice, and she told me what I am doing is all about science. But that saying alone didn't assure me completely.

Can anyone comment on whether it is normal or usual to spend one's times on heuristic works not helpful in developing one's knowledge? or Can anyone elaborate my friend's point?

Yes it is normal. Rome wasn't built in a day and neither will your skills be. Next time you do it, you will probably do it faster and better because you know the software more. Keep trying and eventually it will become second nature. It can be tough and frustrating but you learn through failure. Though you might be able to speed things up if you can find another PhD student will knoweldge with the software, and trade help for beer.

However is it the right software? You might be trying to use the wrong software for your probem. So have you checked what other researchers are using or what your supervisor recommends? I would check that first.

Fortunately, they are the right ones that my supervisor recommended me to use. I see your point, but still need to use time more wisely. Balancing two different activities may help me rather than investigating all the day on a single task.

Are you spending most of your time setting up the simulations or running them? If it is taking forever to run can you get acces to a faster computer? There probably is a high spec desktop somewhere that you can use/book.

Or can you set up batches of simulations to run overnight? A few people in my office write their simulations during the afternoon and leave them to run/complete overnight. Then review their work in the morning, so that you spend your working hours doing constructive tasks.

I would consider using high spec desktop in the office more aggressively. Your suggestion of utilizing bash file to automate a task sounds very helpful. Thank you for advice. Still, the final task - drawing plots and trimming data - that may not be automated are slow and taking long times.
In fact, fundamental point here is maybe what my supervisor had me to do is more heuristic in nature rather than scholary. Not sure about it. I would concillate myself saying that though I am not gaining more knowledge, maybe I am contributong a very small tiny something to this society.