Suppose you have a 2x2 design and you're testing differences between those 4 groups using ANOVA in SPSS.
This is a graph of your data:
After performing ANOVA, there are 6 possible pairwise comparisons between groups that we can perform. These are:
A - C
B - D
A - D
B - C
A - B
C - D
If I want to perform pairwise comparisons, I would usually use this script after the UNIANOVA command:
/EMMEANS=TABLES(Var1*Var2) COMPARE (Var1) ADJ(LSD)
/EMMEANS=TABLES(Var1*Var2) COMPARE (Var2) ADJ(LSD)
However, after running this script, the output only contains 4 of the 6 possible comparisons - there are two pairwise comparisons that are missing, and those are:
A - B
C - D
How can I calculate those comparisons?
EMMEANS in UNIANOVA does not provide all pairwise comparisons among the cells in an interaction like this. There are some other procedures, such as GENLIN, that do offer these, but use large-sample chi-square statistics rather than t or F statistics. In UNIANOVA, you can get these using the LMATRIX subcommand, or you can use some trickery with EMMEANS.
For the trickery with EMMEANS, create a single factor with four levels that index the 2x2 layout of cells, then handle that as a one-way model. The main effect for that is the same as the overall 3 degree of freedom model for the 2x2 layout, and of course EMMEANS with COMPARE works fine on that.
Without creating a new variable, you can use LMATRIX with:
/LMATRIX "(1,1) - (2,2)" var1 1 -1 var2 1 -1 var1*var2 1 0 0 -1
/LMATRIX "(1,2) - (2,1)" var1 1 -1 var1 -1 1 var1*var2 0 1 -1 0
The quoted pieces are labels, indicating the cells in the 2x2 design being compared.
Another trick you can use to make specifying the LMATRIX simpler, but without creating a new variable, is to specify the DESIGN with just the interaction term and suppress the intercept. That makes the parameter estimates just the four cell means:
UNIANOVA Y BY var1 var2
/INTERCEPT=EXCLUDE
/DESIGN var1*var1
/LMATRIX "(1,1) - (2,2)" var1*var2 1 0 0 -1
/LMATRIX "(1,2) - (2,1)" var1*var1 0 1 -1 0.
In this case the one effect shown in the ANOVA table is a 4 df effect testing all means against 0, so it's not of interest, but the comparisons you want are easily obtained. Note that this trick only works with procedures that don't reparameterize to full rank.
Related
I have 5 variables for one questionnaire about social support. I want to define the group with low vs. high support. According to the authors low support is defined as a sum score <= 18 AND two items scoring <= 3.
It would be great to get a dummy variable which shows which people are low vs high in support.
How can I do this in the syntax?
Thanks ;)
Assuming your variables are named Var1, Var2 .... Var5, and that they are consecutive in the dataset, this should work:
recode Var1 to Var5 (1 2 3=1)(4 thr hi=0) into L1 to L5.
compute LowSupport = sum(Var1 to Var5) <= 18 and sum(L1 to L5)>=2.
execute.
New variable LowSupport will have value 1 for rows that have the parameters you defined and 0 for other rows.
Note: If your variables are not consecutive you'll have to list all of them instead of using Var1 to var5.
I am reading the book "Artificial Intelligence" by Stuart Russell and Peter Norvig (Chapter 18). The following paragraph is from the decision trees context.
For a wide variety of problems, the decision tree format yields a
nice, concise result. But some functions cannot be represented
concisely. For example, the majority function, which returns true if
and only if more than half of the inputs are true, requires an
exponentially large decision tree.
In other words, decision trees are good for some kinds of functions
and bad for others. Is there any kind of representation that is
efficient for all kinds of functions? Unfortunately, the answer is no.
We can show this in a general way. Consider the set of all Boolean
functions on "n" attributes. How many different functions are in this
set? This is just the number of different truth tables that we can
write down, because the function is defined by its truth table.
A truth table over "n" attributes has 2^n rows, one for each
combination of values of the attributes.
We can consider the “answer” column of the table as a 2^n-bit number
that defines the function. That means there are (2^(2^n)) different
functions (and there will be more than that number of trees, since
more than one tree can compute the same function). This is a scary
number. For example, with just the ten Boolean attributes of our
restaurant problem there are 2^1024 or about 10^308 different
functions to choose from.
What does author mean by "answer" column of the table as a 2^n-bit number that defines the function?
How did author derive (2^(2^n)) different functions?
Please elaborate on above question, preferably with simple example, such as n = 3.
Consider a general truth table for a 3-input function, where the result for each triple is also a Boolean (1 or 0), represented by variables i through 'p':
A B C f(a,b,c)
0 0 0 i
0 0 1 j
0 1 0 k
0 1 1 l
1 0 0 m
1 0 1 n
1 1 0 o
1 1 1 p
We can now represent any function on three variables as an 8-bit number, ijklmnop. For instance, and is 00000001; or is 01111111; one_hot (exactly one input True) is 01101000.
For 3 variables, you have 2^3 bits in the "answer", the complete function definition. Since there are 8 bits in the "answer", there are 2^8 possible functions we can define.
Does that outline the field of comprehension for you?
More detail on an example function
You simply (once you see the pattern) make the eight bits correspond to the entires in the table. For instance, the table for one-hot looks like this:
A B C f(a,b,c)
0 0 0 0
0 0 1 1
0 1 0 1
0 1 1 0
1 0 0 1
1 0 1 0
1 1 0 0
1 1 1 0
Reading down the "answer" column, labeled f(a,b,c), you get the 8-bit sequence 01101000. That 8-bit number is sufficient to completely define the function: the rows listing all the combinations of a, b, c are in a fixed (numerical) sequence.
You can write any such function in a template format:
def and(a, b, c):
and_def = '00000001'
index = 4*a + 2*b + 1*c
return and_def[index]
Now, if we generalize this to any 3-input binary function:
def_bin_func(a, b, c, func_def)
return func_def[4*a + 2*b + 1*c]
If you wish, you can further generalize the template for a list of inputs: concatenate the bits and use that integer as the index into the func_def string.
Does that clear it up?
I'm trying to solve some huffman coding problems, but I always get different values for the codewords (values not lengths).
for example, if the codeword of character 'c' was 100, in my solution it is 101.
Here is an example:
Character Frequency codeword my solution
A 22 00 10
B 12 100 010
C 24 01 11
D 6 1010 0110
E 27 11 00
F 9 1011 0111
Both solutions have the same length for codewords, and there is no codeword that is prefix of another codeword.
Does this make my solution valid ? or it has to be only 2 solutions, the optimal one and flipping the bits of the optimal one ?
There are 96 possible ways to assign the 0's and 1's to that set of lengths, and all would be perfectly valid, optimal, prefix codes. You have shown two of them.
There exist conventions to define "canonical" Huffman codes which resolve the ambiguity. The value of defining canonical codes is in the transmission of the code from the compressor to the decompressor. As long as both sides know and agree on how to unambiguously assign the 0's and 1's, then only the code length for each symbol needs to be transmitted -- not the codes themselves.
The deflate format starts with zero for the shortest code, and increments up. Within each code length, the codes are ordered by the symbol values, i.e. sorting by symbol. So for your code that canonical Huffman code would be:
A - 00
C - 01
E - 10
B - 110
D - 1110
F - 1111
So there the two bit codes are assigned in the symbol order A, C, E, and similarly, the four bit codes are assigned in the order D, F. Shorter codes are assigned before longer codes.
There is a different and interesting ambiguity that arises in finding the code lengths. Depending on the order of combination of equal frequency nodes, i.e. when you have a choice of more than two lowest frequency nodes, you can actually end up with different sets of code lengths that are exactly equally optimal. Even though the code lengths are different, when you multiply the lengths by the frequencies and add them up, you get exactly the same number of bits for the two different codes.
There again, the different codes are all optimal and equally valid. There are ways to resolve that ambiguity as well at the time the nodes to combine are chosen, where the benefit can be minimizing the depth of the tree. That can reduce the table size for table-driven Huffman decoding.
For example, consider the frequencies A: 2, B: 2, C: 1, D: 1. You first combine C and D to get 2. Then you have A, B, and C+D all with frequency 2. Now you can choose to combine either A and B, or C+D with A or B. This gives two different sets of bit lengths. If you combine A and B, you get lengths: A-2, B-2, C-2, and D-2. If you combine C+D with B, you get A-1, B-2, C-3, D-3. Both are optimal codes, since 2x2 + 2x2 + 1x2 + 1x2 = 2x1 + 2x2 + 1x3 + 1x3 = 12, so both codes use 12 bits to represent those symbols that many times.
The problem is, that there is no problem.
You huffman tree is valid, it also gives the exactly same results after encoding and decoding. Just think if you would build a huffman tree by hand, there are always more ways to combine items with equal (or least difference) value. E.g. if you have A B C (everyone frequency 1), you can at first combine A and B, and the result with C, or at first B and C, and the result with a.
You see, there are more correct ways.
Edit: Even with only one possible way to combine the items by frequency, you can get different results because you can assign 1 for the left or for the right branch, so you would get different (correct) results.
I have to partition a multiset into two sets who sums are equal. For example, given the multiset:
1 3 5 1 3 -1 2 0
I would output the two sets:
1) 1 3 3
2) 5 -1 2 1 0
both of which sum to 7.
I need to do this using Z3 (smt2 input format) and "Linear Arithmetic Logic", which is defined as:
formula : formula /\ formula | (formula) | atom
atom : sum op sum
op : = | <= | <
sum : term | sum + term
term : identifier | constant | constant identifier
I honestly don't know where to begin with this and any advice at all would be appreciated.
Regards.
Here is an idea:
1- Create a 0-1 integer variable c_i for each element. The idea is c_i is zero if element is in the first set, and 1 if it is in the second set. You can accomplish that by saying that 0 <= c_i and c_i <= 1.
2- The sum of the elements in the first set can be written as 1*(1 - c_1) + 3*(1 - c_2) + ... +
3- The sum of the elements in the second set can be written as 1*c1 + 3*c2 + ...
While SMT-Lib2 is quite expressive, it's not the easiest language to program in. Unless you have a hard requirement that you have to code directly in SMTLib2, I'd recommend looking into other languages that have higher-level bindings to SMT solvers. For instance, both Haskell and Scala have libraries that allow you to script SMT solvers at a much higher level. Here's how to solve your problem using the Haskell, for instance: https://gist.github.com/1701881.
The idea is that these libraries allow you to code at a much higher level, and then perform the necessary translation and querying of the SMT solver for you behind the scenes. (If you really need to get your hands onto the SMTLib encoding of your problem, you can use these libraries as well, as they typically come with the necessary API to dump the SMTLib they generate before querying the solver.)
While these libraries may not offer everything that Z3 gives you access to via SMTLib, they are much easier to use for most practical problems of interest.
I'm puzzling over how to map a set of sequences to consecutive integers.
All the sequences follow this rule:
A_0 = 1
A_n >= 1
A_n <= max(A_0 .. A_n-1) + 1
I'm looking for a solution that will be able to, given such a sequence, compute a integer for doing a lookup into a table and given an index into the table, generate the sequence.
Example: for length 3, there are 5 the valid sequences. A fast function for doing the following map (preferably in both direction) would be a good solution
1,1,1 0
1,1,2 1
1,2,1 2
1,2,2 3
1,2,3 4
The point of the exercise is to get a packed table with a 1-1 mapping between valid sequences and cells.
The size of the set in bounded only by the number of unique sequences possible.
I don't know now what the length of the sequence will be but it will be a small, <12, constant known in advance.
I'll get to this sooner or later, but though I'd throw it out for the community to have "fun" with in the meantime.
these are different valid sequences
1,1,2,3,2,1,4
1,1,2,3,1,2,4
1,2,3,4,5,6,7
1,1,1,1,2,3,2
these are not
1,2,2,4
2,
1,1,2,3,5
Related to this
There is a natural sequence indexing, but no so easy to calculate.
Let look for A_n for n>0, since A_0 = 1.
Indexing is done in 2 steps.
Part 1:
Group sequences by places where A_n = max(A_0 .. A_n-1) + 1. Call these places steps.
On steps are consecutive numbers (2,3,4,5,...).
On non-step places we can put numbers from 1 to number of steps with index less than k.
Each group can be represent as binary string where 1 is step and 0 non-step. E.g. 001001010 means group with 112aa3b4c, a<=2, b<=3, c<=4. Because, groups are indexed with binary number there is natural indexing of groups. From 0 to 2^length - 1. Lets call value of group binary representation group order.
Part 2:
Index sequences inside a group. Since groups define step positions, only numbers on non-step positions are variable, and they are variable in defined ranges. With that it is easy to index sequence of given group inside that group, with lexicographical order of variable places.
It is easy to calculate number of sequences in one group. It is number of form 1^i_1 * 2^i_2 * 3^i_3 * ....
Combining:
This gives a 2 part key: <Steps, Group> this then needs to be mapped to the integers. To do that we have to find how many sequences are in groups that have order less than some value. For that, lets first find how many sequences are in groups of given length. That can be computed passing through all groups and summing number of sequences or similar with recurrence. Let T(l, n) be number of sequences of length l (A_0 is omitted ) where maximal value of first element can be n+1. Than holds:
T(l,n) = n*T(l-1,n) + T(l-1,n+1)
T(1,n) = n
Because l + n <= sequence length + 1 there are ~sequence_length^2/2 T(l,n) values, which can be easily calculated.
Next is to calculate number of sequences in groups of order less or equal than given value. That can be done with summing of T(l,n) values. E.g. number of sequences in groups with order <= 1001010 binary, is equal to
T(7,1) + # for 1000000
2^2 * T(4,2) + # for 001000
2^2 * 3 * T(2,3) # for 010
Optimizations:
This will give a mapping but the direct implementation for combining the key parts is >O(1) at best. On the other hand, the Steps portion of the key is small and by computing the range of Groups for each Steps value, a lookup table can reduce this to O(1).
I'm not 100% sure about upper formula, but it should be something like it.
With these remarks and recurrence it is possible to make functions sequence -> index and index -> sequence. But not so trivial :-)
I think hash with out sorting should be the thing.
As A0 always start with 0, may be I think we can think of the sequence as an number with base 12 and use its base 10 as the key for look up. ( Still not sure about this).
This is a python function which can do the job for you assuming you got these values stored in a file and you pass the lines to the function
def valid_lines(lines):
for line in lines:
line = line.split(",")
if line[0] == 1 and line[-1] and line[-1] <= max(line)+1:
yield line
lines = (line for line in open('/tmp/numbers.txt'))
for valid_line in valid_lines(lines):
print valid_line
Given the sequence, I would sort it, then use the hash of the sorted sequence as the index of the table.