In [7]:

    

using DataFrames
X = readtable("../data/chembl_19_mf1/chembl-IC50-10targets.csv", header=true)
rename!(X, [:row, :col], [:compound, :target])


    

    
Out[7]:




compound
target
value
1
1875
1
3516.24
2
1892
1
3484.38
3
1930
1
136.0
4
1981
1
847.3
5
2024
1
7345.73
6
2025
1
1081.87
7
2081
1
1177.99
8
2154
1
699.18
9
2219
1
3254.9
10
2237
1
1690.0

In [8]:

    

X[:, :value] = log10(X[:, :value])


    

    
Out[8]:





10-element DataArray{Float64,1}:
 3.54608
 3.54213
 2.13354
 2.92804
 3.86603
 3.03418
 3.07114
 2.84459
 3.51254
 3.22789

In [9]:

    

idx = sample(1:size(X,1), int(floor(20/100 * size(X,1))); replace=false)


    

    
Out[9]:





2-element Array{Int64,1}:
 8
 7

In [11]:

    

int(floor(20/100 * size(X,1)))


    

    
Out[11]:





2

In [12]:

    

probe_vec = array(X[idx,:])


    

    
Out[12]:





2x3 Array{Real,2}:
 2154  1  2.84459
 2081  1  3.07114

In [13]:

    

ratings_test = probe_vec[:,3] .< log10(200)


    

    
Out[13]:





2-element BitArray{1}:
 false
 false

In [ ]: