Hello,
I am attempting to generate an new array C by using conditional selections from two existing DataFrames (A and B).
The first pre-existing DataFrame (A) is 18 x 18 and contains Float64 values that I would like to assign to specific entries to populate the new array. Their position in the new array depends on values from a different DataFrame.
A=DataFrame(
xDST=[-0.18,-0.16,-0.19,-0.06,-0.23,-0.1,-0.46,-0.5,-0.33,0,-0.04,0.09,0.08,0.09,0.02,0.09,0.35,1],
xK=[0.02,-0.08,-0.01,-0.01,-0.08,-0.16,-0.15,-0.41,-0.5,0.1,0.06,0.18,0.02,0.21,0.19,0.35,1,0.35],
xQB=[0.1,0.09,-0.11,-0.12,-0.05,0,0,-0.15,-0.46,0.35,0.39,0.34,0.16,0.21,0.09,1,0.35,0.09],
xRB1=[0.05,0.03,0,0.06,0.04,-0.31,0,-0.16,-0.1,-0.14,-0.05,-0.07,-0.12,-0.02,1,0.09,0.19,0.02],
xRB2=[0.04,-0.08,-0.05,-0.03,0.03,0.04,-0.05,-0.08,-0.23,0.08,-0.08,0.08,-0.08,1,-0.02,0.21,0.21,0.09],
xTE=[-0.08,0.11,-0.09,-0.1,-0.03,0.06,-0.12,-0.01,-0.06,-0.12,-0.12,-0.08,1,-0.08,-0.12,0.16,0.02,0.08],
xWR1=[-0.01,-0.08,-0.03,-0.09,-0.05,0,-0.11,-0.01,-0.19,-0.02,0.13,1,-0.08,0.08,-0.07,0.34,0.18,0.09],
xWR2=[0.09,-0.06,-0.08,0.11,-0.08,0.03,0.09,-0.08,-0.16,-0.01,1,0.13,-0.12,-0.08,-0.05,0.39,0.06,-0.04],
xWR3=[0.12,0.09,-0.01,-0.08,0.04,0.05,0.1,0.02,-0.18,1,-0.01,-0.02,-0.12,0.08,-0.14,0.35,0.1,0],
yDST=[0,-0.04,0.09,0.08,0.09,0.02,0.09,0.35,1,-0.18,-0.16,-0.19,-0.06,-0.23,-0.1,-0.46,-0.5,-0.33],
yK=[0.1,0.06,0.18,0.02,0.21,0.19,0.35,1,0.35,0.02,-0.08,-0.01,-0.01,-0.08,-0.16,-0.15,-0.41,-0.5],
yQB=[0.35,0.39,0.34,0.16,0.21,0.09,1,0.35,0.09,0.1,0.09,-0.11,-0.12,-0.05,0,0,-0.15,-0.46],
yRB1=[-0.14,-0.05,-0.07,-0.12,-0.02,1,0.09,0.19,0.02,0.05,0.03,0,0.06,0.04,-0.31,0,-0.16,-0.1],
yRB2=[0.08,-0.08,0.08,-0.08,1,-0.02,0.21,0.21,0.09,0.04,-0.08,-0.05,-0.03,0.03,0.04,-0.05,-0.08,-0.23],
yTE=[-0.12,-0.12,-0.08,1,-0.08,-0.12,0.16,0.02,0.08,-0.08,0.11,-0.09,-0.1,-0.03,0.06,-0.12,-0.01,-0.06],
yWR1=[-0.02,0.13,1,-0.08,0.08,-0.07,0.34,0.18,0.09,-0.01,-0.08,-0.03,-0.09,-0.05,0,-0.11,-0.01,-0.19],
yWR2=[-0.01,1,0.13,-0.12,-0.08,-0.05,0.39,0.06,-0.04,0.09,-0.06,-0.08,0.11,-0.08,0.03,0.09,-0.08,-0.16],
yWR3=[1,-0.01,-0.02,-0.12,0.08,-0.14,0.35,0.1,0,0.12,0.09,-0.01,-0.08,0.04,0.05,0.1,0.02,-0.18],
row=["yWR3","yWR2","yWR1","yTE","yRB2","yRB1","yQB","yK","yDST","xWR3","xWR2","xWR1","xTE","xRB2","xRB1","xQB","xK","xDST"])
The second DataFrame is 28 x 7 (B) and contains five “string” type columns (A first (:First) and last (:Last) name column that serve as identifiers, a categorical string column (:Position), a categorical string column (:Team), and a categorical string column (:Opponent), and two Float64 columns (:Projection, and :StdDev).
B = DataFrame(
First = ["Jerry", "Jake", "Timmy", "Jeff", "Jim", "Daniel", "Steven", "Paul", "Ron", "Peter", "Michael", "Ben", "Andy", "Tom", "Jeremy", "Adam", "Blake", "Greg", "William", "Ken", "Charles", "Ed", "Sandy", "Ezekiel", "Jarrod", "Drew", "Alvin", "Shakell"],
Last = ["Green", "Blue", "Tan", "Black", "Little", "Yellow", "Purple", "Walker", "Spears", "Ryan", "Collins", "Smith", "Fitzpatrick", "Byers", "Foster", "Valentine", "Castro", "Hull", "Croft", "Perry", "Franks", "Philips", "Crowe", "Petty", "Pitt", "Kennedy", "Woods", "Barclay"],
Position = ["QB", "QB", "QB", "QB", "RB", "RB", "RB", "WR", "RB", "WR", "WR", "WR", "WR", "WR", "RB", "WR", "RB", "WR", "WR", "RB", "RB", "WR", "RB", "WR", "WR", "WR", "WR", "WR"],
Team = ["Bears", "Lions", "Tigers", "Wolves", "Bears", "Tigers", "Lions", "Tigers", "Wolves", "Lions", "Bears", "Wolves", "Tigers", "Lions", "Wolves", "Bears", "Tigers", "Tigers", "Lions", "Lions", "Bears" , "Tigers", "Wolves", "Wolves", "Bears", "Lions", "Wolves", "Lions"],
Opponent = ["Lions", "Bears", "Wolves", "Tigers", "Lions", "Wolves", "Bears", "Wolves", "Tigers", "Bears", "Lions", "Tigers", "Wolves", "Bears", "Tigers", "Lions", "Wolves", "Wolves", "Bears", "Bears", "Lions", "Wolves", "Tigers", "Tigers", "Lions", "Bears", "Tigers", "Bears"],
Projection = [20.2, 19.7, 18.9, 17.5, 15.3, 15.0, 14.8, 14.75, 13.4, 12.4, 12.14, 11.4, 10.5, 10.2, 9.7, 9.25, 8.95, 8.45, 7.5, 6.2, 5.9, 5.2, 4.6, 3.2, 2.5, 2.3, 1.9, 1.37],
Variance = (rand(Float64, 28) .* [20.2, 19.7, 18.9, 17.5, 15.3, 15.0, 14.8, 14.75, 13.4, 12.4, 12.14, 11.4, 10.5, 10.2, 9.7, 9.25, 8.95, 8.45, 7.5, 6.2, 5.9, 5.2, 4.6, 3.2, 2.5, 2.3, 1.9, 1.37]/2)
)
For all rows with entries from the same :Team, for each position (QB, RB, WR), I want to populate an array C by assigning to the same row index in B to the corresponding positional “x” value from A ( xQB, xRB1, etc ) for their team and opponent. For example, in row 1, Jerry Green is the QB for the Bears (whose opponent is the Lions), so in C[1,1] I want to assign the value for xQB (A column 3) by xQB (A row 16), = 1).
Next, in C[1,2], the next player (Jake Blue) is the QB for the Lions (whose opponent is the Bears), so in C[1,2], I want to assign the value for yQB (A column 12) by xQB (row 16), = 0).
For C[5,1] , we have Jim Little, an RB for the Bears with the highest projected score at 15.3. Because Jim’s projection is the highest for all RBs on his team, he would be the RB1. The next highest projection for the Bears (Charles Franks - projection 5.9) would be their RB2. A similar assignment follows for WRs on the same team. So in C[5,1], I want to assign the value for xRB1 (A column 4) by xQB (A row 16), = 0.09).
C should start to look like this:
8Ă—8 Array{Float64,2}:
1.0 0.0 0.0 0.0 0.09 0.0 0.0 0.0
0.0 1.0 0.0 0.0 0.0 0.0 0.09 0.0
0.0 0.0 1.0 0.0 0.0 0.09 0.0 0.34
0.0 0.0 0.0 1.0 0.0 0.0 0.0 -0.11
0.09 0.0 0.0 0.0 1.0 0.0 -0.31 0.0
0.0 0.0 0.09 0.0 0.0 1.0 0.0 -0.07
0.0 0.09 0.0 0.0 -0.31 0.0 1.0 0.0
0.0 0.0 0.34 -0.11 0.0 -0.07 0.0 1.0
but end up as a 28 x 28. The other wrinkle is that a :Team in B might have more players for a position like RB or WR than entries that correspond in A. For example, in B the Wolves have three RBs, but there are only entries for RB1 and RB2 in A. I would like the additional players be assigned the A entry for the lowest :Projection. So the third highest projected RB (RB3 for the Wolves, Sandy Crowe) would be assigned the RB2 value from A. Similarly for WRs, any additional WRs on a team beyond three would be assigned WR3 value.
Ultimately, I want to use C in the function
cor2cov(C, s)
with s the :StdDev vector to compute a covariance matrix.
What I have tried is:
teams = unique(B[!, :Team])
opponents = B[!, :Opponent]
pair_info = zeros(Float64, size(B))
for num=1:size(B)[1]
if B[!, :Team][num] == teams[1]
if B[!, :Position][num] == "QB"
pair_info[num] = A[3,16]
elseif B[!, :Position][num] == "WR"
WRval = findall(( B[!, :Position] .== "WR") .& (B[!, :Team] .== teams[num] ))
WR_index = [A[12,3], A[11,3], A[10,3], A[10,3], A[10,3]]
for i in eachindex(WRval)
pair_info[WRval[i]] = WR_index[i]
end
elseif B[!, :Position][num] == "RB"
RBval = findall(( B[!, :Position] .== "RB") .& (B[!, :Team] .== teams[num] ))
RB_index = [A[15,3], A[14,3], A[14,3]]
for i in eachindex(RBval)
pair_info[RBval[i]] = RB_index[i]
end
end
end
end
team_pairs = hcat(pair_info)
But I get a method error:
MethodError: no method matching zeros(::Type{Float64}, ::DataFrame)
Closest candidates are:
zeros(::Type{T}, !Matched::Union{Integer, AbstractUnitRange}...) where T at array.jl:500
zeros(::Type{T}, !Matched::Tuple{}) where T at array.jl:509
zeros(::Type{T}, !Matched::Tuple{Vararg{Integer,N}}) where {T, N} at array.jl:504
...
Stacktrace:
[1] top-level scope at In[258]:1
Essentially, I am not quite sure how to structure the conditional if, elseif statements correctly with the findall index code to achieve the results I want. Could someone kindly point me in the right direction? Thanks very much!