Answer:
Step-by-step explanation:
- This is a neat question. I suggest that you graph it on desmos. You will lean a bunch by doing it. These are the graphs you should put on there.
- y = 1/(abs(x - a) - 1):In red
- When you use your down arrow, a new entry will be given you. You can use your mouse to operate the slider you get when you type a = 5
- so your next graph should be a = 5
- Use your down arrow again. Put in x = 4: Blue Line
- Use your down arrow again. Put in x = 6: Green Line
Be patient with this. Play with the sliders to see what they do. I believe you are correct, but I am not certain of the second union although it does look correct.
The graph I have enclosed is of what I have described in the steps above. Make sure you try it all out.
Below are \triangle ABC△ and \triangle DEF△DEFtriangle, D, E, F. We assume that AB=DEAB=DEA, B, equals, D, E, BC=EFBC=EFB, C, eq
Novay_Z [31]
Answer:
- (A) AB = DE and segments with same length are congruent
Step-by-step explanation:
Justification of step 1 is based on the congruence of segments AB and DE.
<u>Therefore correct choice is the first one:</u>
- (A) AB = DE and segments with same length are congruent
It’s 10. This is because 27/3 is 9 so the ratio is by 9
Answer:
The experimental probability of rolling an odd number is
55%, which is
5% more than the theoretical probability.
Step-by-step explanation:
3 4 5 2 7 1 3 7 2 6 2 1 7 3 6 1 8 3 5 6
The odd outcomes are in bold below:
3 4 5 2 7 1 3 7 2 6 2 1 7 3 6 1 8 3 5 6
There were 20 rolls of the die. 11 of the 20 rolls were odd.
experimental probability of rolling an odd number = 11/20 = 55%
The die has 4 even numbers and 4 odd numbers. The theoretical probability of rolling an odd number is
theoretical probability of rolling an odd number = 4/8 = 1/2 = 50%
The experimental probability is 5% more than the theoretical probability.
Answer: The experimental probability of rolling an odd number is
55%, which is
5% more than the theoretical probability.
Yes, they certainly do exist.
Do you have a question to ask ?