Statistics for "Computer and mathematical occupations" covers such a wide variety of occupations as to be meaningless in the context of your own claims about "junior dev" positions. But for the sake of argument..
According to your source, the unemployment rate for women in Computer and mathematical occupations in May was 2.6 vs 1.8 percent for men. That's actually a 31 percent higher amount. Still high but not 50 as you claimed in your previous post. It also says the rate for May of 2015 was exactly the same for both genders at 1.5 percent. This issue has gotten a lot of press in the last 12 months so it is very unlikely that suddenly the numbers are disparate as a result of discrimination after they were the same a year ago.
And in other sectors on your chart like in Management, professional, and related occupations, the unemployment figures are the opposite. Women unemployment went from higher to lower than the men's rate. I mean, in Architecture and engineering occupations, the unemployment rate of men is literally 100 percent higher than women at 2.2 vs 1.1 percent. It just looks like natural statistical gyrations of numbers and not conclusive or even particularly indicative of anything.
And the point is, if you are saying that huge numbers of women are graduating with these degrees and then being shut out of the job market, you can't just stop there, you have to back that up. Even ignoring the fact that just because someone has a degree doesn't mean they're qualified, quoting the percent of females in entry level dev jobs doesn't say anything about where the other women went or what they are doing. I'm pointing out that if they're out there looking for these jobs then the unemployment statistics should reflect that and via your source, they don't. Other posters have suggested graduate school and other jobs not described as entry level dev and so far that explanation seems more plausible. Ask yourself, what are CS graduates qualified to do besides be junior devs and you may be surprised with how many answers there are.
I think where your confusion lies is in order for there to be discrimination, there has to actually be discrimination. Point out the discrimination, who's doing it, which companies, what policy allows this, and we can all rally and do something about it. But just throwing out cherry picked employment numbers and claiming bias is somehow self evident which is what your argument implies isn't good enough. It actually comes across as a kind of weird p-hacking[0]. Even worse is when you introduce a credibility problem by conflating different employment sectors and inflating your numbers like you have here.
[0]https://en.wikipedia.org/wiki/Data_dredging