Hi guys,

Given cquantile(Chisq(1), .005), we can get its Chi square table value 7.8

Is there anyway to get the corresponding p value when given Chi square table value and freedom?

Thanks,

Hongwei

Hi guys,

Given cquantile(Chisq(1), .005), we can get its Chi square table value 7.8

Is there anyway to get the corresponding p value when given Chi square table value and freedom?

Thanks,

Hongwei

```
using Distributions
ccdf(Chisq(1), 7.8)
```

2 Likes

That one I can help with:

```
julia> 1-cdf(Chisq(1),7.8)
0.00522462344006136
```

Thanks a lot, Tamas!

Thanks a lot, Mcreel! It seems ccdf and cdf are complementary of each other.

`ccdf`

may offer more precision in some cases. Computing `1 - cdf`

can be numerically problematic if `cdf`

is very close to 1:

julia> (1 - 0.9999999999999999) + 0.0000000000000001

2.1102230246251565e-16julia> 1 - (0.9999999999999999 + 0.0000000000000001)

0.0

2 Likes

Looking at this again, I wrote this example way too quickly and totally screwed it up by putting a plus sign instead of a minus sign in the first computation. The relevant comparison would be something like this:

julia> 1 - (0.9999999999999999 + 0.0000000000000001)

0.0julia> (1 - 0.9999999999999999) - 0.0000000000000001

1.1022302462515656e-17

1 Like