Supported Optional Parameters#
Note
This page lists optional parameters for C APIs only.
In all the following tables, \(\varepsilon\), refers to a safe machine precision (twice the actual machine precision) for the given floating point data type.
Linear models#
The following options are supported.
Option name |
Type |
Default |
Description |
Constraints |
|---|---|---|---|---|
print level |
integer |
\(i=0\) |
Set level of verbosity for the solver. |
\(0 \le i \le 5\) |
optim convergence tol |
real |
\(r=10/2\sqrt{2\,\varepsilon}\) |
Tolerance to declare convergence for the iterative optimization step. See option in the corresponding optimization solver documentation. |
\(0 < r < 1\) |
intercept |
integer |
\(i=0\) |
Add intercept variable to the model. |
\(0 \le i \le 1\) |
optim method |
string |
\(s=\) auto |
Select optimization method to use. |
\(s=\) auto, bfgs, cg, chol, cholesky, coord, lbfgs, lbfgsb, qr, sparse_cg, or svd. |
optim iteration limit |
integer |
\(i=10000\) |
Maximum number of iterations to perform in the optimization phase. Valid only for iterative solvers, e.g. L-BFGS-B, Coordinate Descent, etc. |
\(1 \le i\) |
scaling |
string |
\(s=\) auto |
Scale or standardize feature matrix and response vector. Matrix is copied and then rescaled. Option key value auto indicates that rescaling type is chosen by the solver (this also includes no scaling). |
\(s=\) auto, centering, no, none, scale, scale only, standardise, or standardize. |
print options |
string |
\(s=\) no |
Print options. |
\(s=\) no, or yes. |
optim coord skip min |
integer |
\(i=2\) |
Minimum times a coordinate change is smaller than coord skip tol to start skipping. |
\(2 \le i\) |
optim coord skip max |
integer |
\(i=100\) |
Maximum times a coordinate can be skipped, after this the coordinate is checked. |
\(10 \le i\) |
debug |
integer |
\(i=0\) |
Set debug level (internal use). |
\(0 \le i \le 3\) |
logistic constraint |
string |
\(s=\) ssc |
Affects only multinomial logistic regression. Type of constraint put on coefficients. This will affect number of coefficients returned. RSC - means we choose a reference category whose coefficients will be set to all 0. This results in K-1 class coefficients for problems with K classes. SSC - means the sum of coefficients class-wise for each feature is 0. It will result in K class coefficients for problems with K classes. |
\(s=\) reference category, rsc, ssc, symmetric, or symmetric side. |
optim time limit |
real |
\(r=10^6\) |
Maximum time limit (in seconds). Solver will exit with a warning after this limit. Valid only for iterative solvers, e.g. L-BFGS-B, Coordinate Descent, etc. |
\(0 < r\) |
lambda |
real |
\(r=0\) |
Penalty coefficient for the regularization terms: lambda( (1-alpha)/2 L2 + alpha L1 ). |
\(0 \le r\) |
alpha |
real |
\(r=0\) |
Coefficient of alpha in the regularization terms: lambda( (1-alpha)/2 L2 + alpha L1 ). |
\(0 \le r \le 1\) |
optim progress factor |
real |
\(r=\frac{10}{\sqrt{2\,\varepsilon}}\) |
Factor used to detect convergence of the iterative optimization step. See option in the corresponding optimization solver documentation. |
\(0 \le r\) |
Principal component analysis#
The following options are supported.
Option name |
Type |
Default |
Description |
Constraints |
|---|---|---|---|---|
svd solver |
string |
\(s=\) auto |
Which LAPACK routine to use for the underlying singular value decomposition. |
\(s=\) auto, gesdd, gesvd, gesvdx, or syevd. |
degrees of freedom |
string |
\(s=\) unbiased |
Whether to use biased or unbiased estimators for standard deviations and variances. |
\(s=\) biased, or unbiased. |
pca method |
string |
\(s=\) covariance |
Compute PCA based on the covariance or correlation matrix. |
\(s=\) correlation, covariance, or svd. |
store u |
integer |
\(i=0\) |
Whether or not to store the matrix U from the SVD. |
\(0 \le i \le 1\) |
n_components |
integer |
\(i=1\) |
Number of principal components to compute. If 0, then all components will be kept. |
\(0 \le i\) |
k-means clustering#
The following options are supported.
Option name |
Type |
Default |
Description |
Constraints |
|---|---|---|---|---|
algorithm |
string |
\(s=\) lloyd |
Choice of underlying k-means algorithm. |
\(s=\) elkan, hartigan-wong, lloyd, or macqueen. |
initialization method |
string |
\(s=\) random |
How to determine the initial cluster centres. |
\(s=\) k-means++, random, random partitions, or supplied. |
convergence tolerance |
real |
\(r=10^{-4}\) |
Convergence tolerance. |
\(0 \le r\) |
seed |
integer |
\(i=0\) |
Seed for random number generation; set to -1 for non-deterministic results. |
\(-1 \le i\) |
max_iter |
integer |
\(i=300\) |
Maximum number of iterations. |
\(1 \le i\) |
n_init |
integer |
\(i=10\) |
Number of runs with different random seeds (ignored if you have specified initial cluster centres). |
\(1 \le i\) |
n_clusters |
integer |
\(i=1\) |
Number of clusters required. |
\(1 \le i\) |
Decision trees#
The following options are supported.
Option name |
Type |
Default |
Description |
Constraints |
|---|---|---|---|---|
minimum split improvement |
real |
\(r=0.03\) |
Minimum score improvement needed to consider a split from the parent node. |
\(0 \le r\) |
minimum split score |
real |
\(r=0.03\) |
Minimum score needed for a node to be considered for splitting. |
\(0 \le r \le 1\) |
tree building order |
string |
\(s=\) depth first |
Select in which order to explore the nodes. |
\(s=\) breadth first, or depth first. |
feature threshold |
real |
\(r=1e-06\) |
Minimum difference in feature value required for splitting. |
\(0 \le r\) |
maximum features |
integer |
\(i=0\) |
Set the number of features to consider when splitting a node. 0 means take all the features. |
\(0 \le i\) |
print timings |
string |
\(s=\) no |
Print the timings of different parts of the fitting process. |
\(s=\) no, or yes. |
seed |
integer |
\(i=-1\) |
Set the random seed for the random number generator. If the value is -1, a random seed is automatically generated. In this case the resulting classification will create non-reproducible results. |
\(-1 \le i\) |
maximum depth |
integer |
\(i=29\) |
Set the maximum depth of trees. |
\(0 \le i \le 29\) |
predict probabilities |
integer |
\(i=1\) |
evaluate class probabilities (in addition to class predictions).Needs to be 1 if calls to predict_proba or predict_log_probaare made after fit. |
\(0 \le i \le 1\) |
node minimum samples |
integer |
\(i=2\) |
The minimum number of samples required to split an internal node. |
\(2 \le i\) |
scoring function |
string |
\(s=\) gini |
Select scoring function to use. |
\(s=\) cross-entropy, entropy, gini, misclass, misclassification, or misclassification-error. |
Decision forests#
The following options are supported.
Option name |
Type |
Default |
Description |
Constraints |
|---|---|---|---|---|
block size |
integer |
\(i=256\) |
Set the size of the blocks for parallel computations. |
\(1 \le i \le 2147483647\) |
node minimum samples |
integer |
\(i=2\) |
Minimum number of samples to consider a node for splitting. |
\(2 \le i\) |
scoring function |
string |
\(s=\) gini |
Select scoring function to use. |
\(s=\) cross-entropy, entropy, gini, misclass, misclassification, or misclassification-error. |
maximum depth |
integer |
\(i=29\) |
Set the maximum depth of trees. |
\(0 \le i \le 29\) |
seed |
integer |
\(i=-1\) |
Set random seed for the random number generator. If the value is -1, a random seed is automatically generated. In this case the resulting classification will create non-reproducible results. |
\(-1 \le i\) |
tree building order |
string |
\(s=\) depth first |
Select in which order to explore the nodes. |
\(s=\) breadth first, or depth first. |
feature threshold |
real |
\(r=1e-06\) |
Minimum difference in feature value required for splitting. |
\(0 \le r\) |
bootstrap |
string |
\(s=\) yes |
Select whether to bootstrap the samples in the trees. |
\(s=\) no, or yes. |
bootstrap samples factor |
real |
\(r=0.8\) |
Proportion of samples to draw from the data set to build each tree if ‘bootstrap’ was set to ‘yes’. |
\(0 < r \le 1\) |
features selection |
string |
\(s=\) sqrt |
Select how many features to use for each split. |
\(s=\) all, custom, log2, or sqrt. |
number of trees |
integer |
\(i=100\) |
Set the number of trees to compute. |
\(1 \le i\) |
minimum split score |
real |
\(r=0.03\) |
Minimum score needed for a node to be considered for splitting. |
\(0 \le r \le 1\) |
maximum features |
integer |
\(i=0\) |
Set the number of features to consider when splitting a node. 0 means take all the features. |
\(0 \le i\) |
minimum split improvement |
real |
\(r=0.03\) |
Minimum score improvement needed to consider a split from the parent node. |
\(0 \le r\) |
k-nearest neighbors#
The following options are supported.
Option name |
Type |
Default |
Description |
Constraints |
|---|---|---|---|---|
weights |
string |
\(s=\) uniform |
Weight function used to compute the k-nearest neighbors. |
\(s=\) distance, or uniform. |
metric |
string |
\(s=\) euclidean |
Metric used to compute the pairwise distance matrix. |
\(s=\) euclidean, or sqeuclidean. |
algorithm |
string |
\(s=\) brute |
Algorithm used to compute the k-nearest neighbors. |
\(s=\) brute. |
number of neighbors |
integer |
\(i=5\) |
Number of neighbors considered for k-nearest neighbors. |
\(1 \le i\) |
Datastore handle da_datastore#
The following options are supported.
Option name |
Type |
Default |
Description |
Constraints |
|---|---|---|---|---|
csv integers as floats |
integer |
\(i=0\) |
Whether or not to interpret integers as floating point numbers when using autodetection. |
\(0 \le i \le 1\) |
csv datastore precision |
string |
\(s=\) double |
The precision used when reading floating point numbers using autodetection. |
\(s=\) double, or single. |
csv use header row |
integer |
\(i=0\) |
Whether or not to interpret the first row as a header. |
\(0 \le i \le 1\) |
csv warn for missing data |
integer |
\(i=0\) |
If set to 0, return error if missing data is encountered; if set to 1, issue a warning and store missing data as either a NaN (for floating point data) or the maximum value of the integer type being used. |
\(0 \le i \le 1\) |
csv skip footer |
integer |
\(i=0\) |
Whether or not to ignore the last line when reading a CSV file. |
\(0 \le i \le 1\) |
csv delimiter |
string |
\(s=\) , |
The delimiter used when reading CSV files. |
|
csv whitespace delimiter |
integer |
\(i=0\) |
Whether or not to use whitespace as the delimiter when reading CSV files. |
\(0 \le i \le 1\) |
csv decimal |
string |
\(s=\) . |
The character used to denote a decimal point in CSV files. |
|
csv skip initial space |
integer |
\(i=0\) |
Whether or not to ignore initial spaces in CSV file lines. |
\(0 \le i \le 1\) |
csv line terminator |
string |
empty |
The character used to denote line termination in CSV files (leave this empty to use the default). |
|
csv row start |
integer |
\(i=0\) |
Ignore the specified number of lines from the top of the file (note that line numbers in CSV files start at 1). |
\(0 \le i\) |
csv comment |
string |
\(s=\) # |
The character used to denote comments in CSV files (note, if a line in a CSV file is to be interpreted as only containing a comment, the comment character should be the first character on the line). |
|
csv quote character |
string |
\(s=\) “ |
The character used to denote quotations in CSV files. |
|
csv scientific notation character |
string |
\(s=\) e |
The character used to denote powers of 10 in floating point values in CSV files. |
|
csv escape character |
string |
\(s=\) \ |
The escape character in CSV files. |
|
csv thousands |
string |
empty |
The character used to separate thousands when reading numeric values in CSV files. |
|
csv skip rows |
string |
empty |
A comma- or space-separated list of rows to ignore in CSV files. |
|
csv datatype |
string |
\(s=\) auto |
If a CSV file is known to be of a single datatype, set this option to disable autodetection and make reading the file quicker. |
\(s=\) auto, boolean, double, float, integer, or string. |
csv data storage |
string |
\(s=\) column major |
Whether to store data from CSV files in row or column major format. |
\(s=\) column major, or row major. |
csv skip empty lines |
integer |
\(i=0\) |
Whether or not to ignore empty lines in CSV files (note that caution should be used when using this in conjunction with options such as CSV skip rows since line numbers may no longer correspond to the original line numbers in the CSV file). |
\(0 \le i \le 1\) |
csv double quote |
integer |
\(i=0\) |
Whether or not to interpret two consecutive quotechar characters within a field as a single quotechar character. |
\(0 \le i \le 1\) |