Merge branch 'main' into line-search

timpute/figures/__init__.py

@@ -0,0 +1,9 @@
+from ..method_ALS import perform_ALS
+from ..method_CLS import perform_CLS
+from ..method_DO import perform_DO
+
+METHODS = (perform_DO, perform_ALS, perform_CLS)
+METHODNAMES = ["DO","ALS","CLS"]
+SAVENAMES = ["zohar", "alter", "hms", "coh_response"]
+DATANAMES = ['Covid serology', 'HIV serology', 'DyeDrop profiling', 'BC cytokine']
+DROPS = (0.01, 0.05, 0.1, 0.2, 0.3, 0.4, 0.5)

timpute/common.py → timpute/figures/common.py

@@ -55,3 +55,18 @@ def subplotLabel(axs):
     """ Place subplot labels on figure. """
     for ii, ax in enumerate(axs):
         ax.text(-0.2, 1.2, ascii_lowercase[ii], transform=ax.transAxes, fontsize=16, fontweight="bold", va="top")
+
+
+def set_boxplot_color(bp, color):
+    plt.setp(bp['boxes'], color=color)
+    plt.setp(bp['whiskers'], color=color)
+    plt.setp(bp['caps'], color=color)
+    plt.setp(bp['medians'], color=color)
+
+
+def rgbs(color = 0, transparency = None):
+    color_rgbs = [sns.color_palette("bright")[0],sns.color_palette("bright")[8],sns.color_palette("bright")[3]]
+    if transparency is not None:
+        return tuple(list(color_rgbs[color]) + [transparency])
+    else:
+        return color_rgbs[color]

timpute/figures/figure2.py

@@ -1,8 +1,8 @@
-import numpy as np
-from .figure_helper import *
-from ..plot import *
-from ..common import *
 import math
+import numpy as np
+from .figure_helper import loadImputation
+from .common import getSetup, subplotLabel, rgbs
+from figures import METHODS, METHODNAMES, SAVENAMES, DATANAMES
 
 # poetry run python -m timpute.figures.figure2
 

timpute/figures/figure3.py

@@ -1,8 +1,9 @@
-import numpy as np
-from .runImputation import *
-from ..plot import *
-from ..common import *
 import math
+import numpy as np
+from matplotlib.lines import Line2D
+from .figure_helper import loadImputation
+from .common import getSetup, subplotLabel, rgbs, set_boxplot_color
+from figures import METHODS, METHODNAMES, SAVENAMES, DATANAMES, DROPS
 
 # poetry run python -m timpute.figures.figure3
 

timpute/figures/figure4.py

@@ -1,8 +1,7 @@
 import numpy as np
-from .runImputation import *
-from ..plot import *
-from ..common import *
-import math
+from .figure_helper import loadImputation
+from .common import getSetup, subplotLabel, rgbs
+from figures import METHODS, METHODNAMES, SAVENAMES, DATANAMES
 
 # poetry run python -m timpute.figures.figure4
 

timpute/figures/figure5.py

@@ -1,8 +1,8 @@
 import numpy as np
 from .figure_data import bestComps
-from .figure_helper import *
-from ..plot import *
-from ..common import *
+from .figure_helper import loadImputation
+from .common import getSetup, subplotLabel, rgbs
+from figures import METHODS, METHODNAMES, SAVENAMES, DATANAMES
 # from matplotlib.legend_handler import HandlerErrorbar
 
 # poetry run python -m timpute.figures.figure5

timpute/figures/figure_data.py

@@ -1,11 +1,8 @@
-import numpy as np
 import os
-
-from .figure_helper import *
+import numpy as np
+from .figure_helper import runImputation, loadImputation
 from ..generateTensor import generateTensor
-from ..plot import *
-from ..common import *
-import pickle
+from figures import METHODS, METHODNAMES, SAVENAMES, DROPS
 
 # poetry run python -m timpute.figures.figure_data  
 

timpute/figures/figure_helper.py

@@ -3,16 +3,6 @@
 from ..decomposition import Decomposition
 from ..tracker import Tracker
 
-from ..method_ALS import perform_ALS
-from ..method_CLS import perform_CLS
-from ..method_DO import perform_DO
-METHODS = (perform_DO, perform_ALS, perform_CLS)
-METHODNAMES = ["DO","ALS","CLS"]
-SAVENAMES = ["zohar", "alter", "hms", "coh_response"]
-DATANAMES = ['Covid serology', 'HIV serology', 'DyeDrop profiling', 'BC cytokine']
-LINESTYLES = ('dashdot', (0,(1,1)), 'solid', (3,(3,1,1,1,1,1)), 'dotted', (0,(5,1)))
-DROPS = (0.01, 0.05, 0.1, 0.2, 0.3, 0.4, 0.5)
-
 
 def runImputation(data:np.ndarray,
                   max_rr:int,

timpute/figures/memUsage.py

@@ -1,9 +1,13 @@
 # import psutil
+import pickle
+import os
+import numpy as np
 import resource
 import argparse
-from .figure_helper import *
+
+from figures import METHODS, METHODNAMES, SAVENAMES
+from ..decomposition import Decomposition
 from ..generateTensor import generateTensor
-import pickle
 
 # poetry run python -m timpute.figures.dataUsage
 

timpute/figures/supplements.py

@@ -1,8 +1,9 @@
 import pickle
-from ..plot import *
-from ..common import * 
+from figures import METHODNAMES, SAVENAMES, DATANAMES, DROPS
+import numpy as np
 import pandas as pd
-from .figure_helper import *
+from .common import getSetup, rgbs
+from ..decomposition import Decomposition
 
 # poetry run python -m timpute.figures.supplements
 

timpute/generateTensor.py

@@ -8,7 +8,7 @@
 from tensordata.atyeo import data as atyeo
 from tensordata.zohar import data as zohar
 from tensordata.alter import data as alter
-from .import_hmsData import hms_tensor
+from .data.import_hmsData import hms_tensor
 
 def generateTensor(type=None, r=6, shape=(10,10,10), scale=2, distribution='gamma', par=2, missingness=0.0, noise_scale=50):
     """

timpute/initialization.py

@@ -2,7 +2,108 @@
 import tensorly as tl
 from tensorly.tenalg import svd_interface
 from tensorly.random import random_cp
-from .SVD_impute import IterativeSVD
+
+class IterativeSVD(object):
+    def __init__(
+            self,
+            rank,
+            convergence_threshold=1e-7,
+            max_iters=500,
+            random_state=None,
+            min_value=None,
+            max_value=None,
+            verbose=False):
+        self.min_value=min_value
+        self.max_value=max_value
+        self.rank = rank
+        self.max_iters = max_iters
+        self.convergence_threshold = convergence_threshold
+        self.verbose = verbose
+        self.random_state = random_state
+
+    def clip(self, X):
+        """
+        Clip values to fall within any global or column-wise min/max constraints
+        """
+        X = np.asarray(X)
+        if self.min_value is not None:
+            X[X < self.min_value] = self.min_value
+        if self.max_value is not None:
+            X[X > self.max_value] = self.max_value
+        return X
+
+    def prepare_input_data(self, X):
+        """
+        Check to make sure that the input matrix and its mask of missing
+        values are valid. Returns X and missing mask.
+        """
+        if X.dtype != "f" and X.dtype != "d":
+            X = X.astype(float)
+
+        assert X.ndim == 2
+        missing_mask = np.isnan(X)
+        assert not missing_mask.all()
+        return X, missing_mask
+
+    def fit_transform(self, X, y=None):
+        """
+        Fit the imputer and then transform input `X`
+        Note: all imputations should have a `fit_transform` method,
+        but only some (like IterativeImputer in sklearn) also support inductive
+        mode using `fit` or `fit_transform` on `X_train` and then `transform`
+        on new `X_test`.
+        """
+        X_original, missing_mask = self.prepare_input_data(X)
+        observed_mask = ~missing_mask
+        X_filled = X_original.copy()
+        X_filled[missing_mask] = 0.0
+        assert isinstance(X_filled, np.ndarray)
+        X_result = self.solve(X_filled, missing_mask)
+        assert isinstance(X_result, np.ndarray)
+        X_result = self.clip(np.asarray(X_result))
+        X_result[observed_mask] = X_original[observed_mask]
+        return X_result
+
+    def _converged(self, X_old, X_new, missing_mask):
+        F32PREC = np.finfo(np.float32).eps
+        # check for convergence
+        old_missing_values = X_old[missing_mask]
+        new_missing_values = X_new[missing_mask]
+        difference = old_missing_values - new_missing_values
+        ssd = np.sum(difference ** 2)
+        old_norm_squared = (old_missing_values ** 2).sum()
+        # edge cases
+        if old_norm_squared == 0 or \
+                (old_norm_squared < F32PREC and ssd > F32PREC):
+            return False
+        else:
+            return (ssd / old_norm_squared) < self.convergence_threshold
+
+    def solve(self, X, missing_mask):
+        observed_mask = ~missing_mask
+        X_filled = X
+        for i in range(self.max_iters):
+            curr_rank = self.rank
+            self.U, S, V = svd_interface(matrix=X_filled, n_eigenvecs=curr_rank, random_state=self.random_state)
+            X_reconstructed = self.U @ np.diag(S) @ V
+            X_reconstructed = self.clip(X_reconstructed)
+
+            # Masked mae
+            mae = np.mean(np.abs(X[observed_mask] - X_reconstructed[observed_mask]))
+
+            if self.verbose:
+                print(
+                    "[IterativeSVD] Iter %d: observed MAE=%0.6f" % (
+                        i + 1, mae))
+            converged = self._converged(
+                X_old=X_filled,
+                X_new=X_reconstructed,
+                missing_mask=missing_mask)
+            X_filled[missing_mask] = X_reconstructed[missing_mask]
+            if converged:
+                break
+        return X_filled
+
 
 def initialize_fac(tensor: np.ndarray, rank: int, method='svd'):
     """Initialize factors used in `parafac`.