Skip to content

scikit fda map

Carlos Ramos Carreño edited this page May 7, 2019 · 25 revisions

Map of the status of the features in scikit-fda:

Alt text

custom_mark10

digraph G { skfda -> representation skfda -> preprocessing skfda -> datasets skfda -> ml skfda -> exploratory skfda -> inference skfda -> "time series" skfda -> multivariate skfda -> "space temporal"

dense [style=filled,color=lightgrey,URL="https://fda.readthedocs.io/en/latest/modules/autosummary/skfda.grid.FDataGrid.html#skfda.grid.FDataGrid"]
representation -> dense
representation -> incomplete
incomplete -> sparse
incomplete -> longitudinal
basis [style=filled,color=lightgrey,URL="https://fda.readthedocs.io/en/latest/modules/autosummary/skfda.FDataBasis.html#skfda.FDataBasis"]
representation -> basis
Fourier [style=filled,color=lightgrey,URL="https://fda.readthedocs.io/en/latest/modules/autosummary/skfda.basis.Fourier.html#skfda.basis.Fourier"]
basis -> Fourier
BSpline [style=filled,color=lightgrey,URL="https://fda.readthedocs.io/en/latest/modules/autosummary/skfda.basis.BSpline.html#skfda.basis.BSpline"]
basis -> BSpline
monomial [style=filled,color=lightgrey,URL="https://fda.readthedocs.io/en/latest/modules/autosummary/skfda.basis.Monomial.html#skfda.basis.Monomial"]
basis -> monomial
basis -> wavelet

preprocessing -> derivatives
"symmetric difference" [style=filled,color=lightgrey,label="symmetric difference (derivative method, 1d)"]
derivatives -> "symmetric difference"

preprocessing -> registration
"shift registration" [style=filled,color=lightgrey,URL="https://fda.readthedocs.io/en/latest/modules/autosummary/skfda.registration.shift_registration.html#skfda.registration.shift_registration"]
registration -> "shift registration"
"landmark shift" [style=filled,color=lightgrey,URL="https://fda.readthedocs.io/en/latest/modules/autosummary/skfda.registration.landmark_shift.html#skfda.registration.landmark_shift"]
registration -> "landmark shift"
"landmark registration" [style=filled,color=lightgrey,URL="https://fda.readthedocs.io/en/latest/modules/autosummary/skfda.registration.landmark_registration.html#skfda.registration.landmark_registration"]
registration -> "landmark registration"
"elastic registration" [style=filled,color=lightgrey,URL="https://fda.readthedocs.io/en/latest/modules/autosummary/skfda.registration.elastic_registration.html#skfda.registration.elastic_registration"]
registration -> "elastic registration"
"MSE decomposition" [style=filled,color=lightgrey,URL="https://fda.readthedocs.io/en/latest/modules/autosummary/skfda.registration.mse_decomposition.html#skfda.registration.mse_decomposition"]
registration -> "MSE decomposition"

preprocessing -> smoothing
smoothing -> kernel

exploratory -> "dimensionality reduction"
exploratory -> visualization
exploratory -> depth
exploratory -> metrics
exploratory -> outliers
exploratory -> stats

FM [style=filled,color=lightgrey,URL="https://fda.readthedocs.io/en/latest/modules/autosummary/skfda.depth_measures.fraiman_muniz_depth.html#skfda.depth_measures.fraiman_muniz_depth"]
depth -> FM
BD [style=filled,color=lightgrey,URL="https://fda.readthedocs.io/en/latest/modules/autosummary/skfda.depth_measures.band_depth.html#skfda.depth_measures.band_depth"]
depth -> BD
MBD [style=filled,color=lightgrey,URL="https://fda.readthedocs.io/en/latest/modules/autosummary/skfda.depth_measures.modified_band_depth.html#skfda.depth_measures.modified_band_depth"]
depth -> MBD
depth -> "h-mode"
depth -> "random projections"
depth -> median
boxplot [style=filled,color=lightgrey,URL="https://fda.readthedocs.io/en/latest/modules/fdata_boxplot.html"]
depth -> boxplot
depth -> "depth outliers"
depth -> "DD plot"

"dimensionality reduction" -> projection
"dimensionality reduction" -> "variable selection"

projection -> FPCA
projection -> FPLS

"variable selection" -> RKHS
"variable selection" -> MH
"variable selection" -> RMH
"variable selection" -> mRMR
"variable selection" -> wrapper
wrapper -> Fwd
wrapper -> Bwd

Lp [style=filled,color=lightgrey,URL="https://fda.readthedocs.io/en/latest/modules/autosummary/skfda.metrics.lp_distance.html#skfda.metrics.lp_distance"]
metrics -> Lp
Linf [label="L∞"]
metrics -> Linf

outliers -> "MS plot"
outliers -> outliergram

visualization -> graphs
visualization -> boxplot
visualization -> "DD plot"
visualization -> "MS plot"
visualization -> outliergram

stats -> mean
stats -> median
stats -> std
stats -> gmean
stats -> trimmed
stats -> robust

inference -> intervals
inference -> tests

intervals -> bootstrap

meanTest [label="mean"]
tests -> meanTest
tests -> ANOVA
tests -> homogeneity

ml -> clustering
ml -> regression
ml -> classification

clustering -> "K-means"
clustering -> "Fuzzy K-means"
clustering -> hierarchical

regression -> linear
regression -> nonparametric
regression -> GLM
regression -> PCA
regression -> PLS
regression -> logistic
regression -> penalized
KnnRegression [label="K-nn"]
regression -> KnnRegression

KnnClass [label="K-nn"]
classification -> KnnClass
classification -> centroid

}

custom_mark10

The map is also being done as an Euler diagram: Alt text

euler_all digraph G {

subgraph cluster_skfda {

 style=rounded
 label="scikit-fda";
 fontsize = 40;
 
 subgraph cluster_top {
     style=invis
     fontsize = 20;
   
     subgraph cluster_exploratory {
       style=rounded
       label="exploratory";
       
       outliers[shape=box,style=rounded]
       dimensionality[shape=box,style=rounded,label="dimensionality reduction"]
       depth[shape=box,style=rounded]
       visualization[shape=box,style=rounded]
       statistics[shape=box,style=rounded]
     }
     subgraph cluster_representation {
       style="rounded"
       label="representation";
       
       basis[shape=box,style="rounded"]
       incomplete[shape=box,style="rounded"]
       dense[shape=box,style="rounded"]
     }
 }
 
 subgraph cluster_bottom {
     style=invis
     fontsize = 20;

     subgraph cluster_ml {
       style=rounded
       label="machine learning";
       
       clustering[shape=box,style=rounded]
       classification[shape=box,style=rounded]
       regression[shape=box,style=rounded]
     }
     subgraph cluster_inference {
       style=rounded
       label="inference";
       
       intervals[shape=box,style=rounded]
       tests[shape=box,style=rounded]
     }
     subgraph cluster_preprocessing {
       style=rounded
       label="preprocessing";
       
       derivatives[shape=box,style=rounded]
       distances[shape=box,style=rounded]
       registration[shape=box,style=rounded]
       smoothing[shape=box,style=rounded]
     }
 }
 
 dense -> smoothing[style = invis]

} } euler_all

Clone this wiki locally