testing files and modified defaults to absolute paths

.gitignore

@@ -112,3 +112,21 @@ output/*
 cache
 
 .idea/
+JamesFig2.png
+JamesFig4.png
+JamesFig5.png
+JamesFig6.png
+JamesFigA1.png
+JamesFigA2.png
+JamesFigS1.png
+JamesIterfigure2.png
+JamesIterfigure4.png
+JamesIterfigure5.png
+JamesIterfigure6.png
+JamesIterfigureA1.png
+JamesIterfigureA2.png
+JamesIterfigureS1.png
+JamesIterfigureS2.png
+JamesIterfigureS3.png
+JamesIterfigureS4.png
+JamesIterfigureS5.png

tfac/dataImport.py

@@ -7,7 +7,7 @@
 import scipy.cluster.hierarchy as sch
 from sklearn.preprocessing import scale
 
-from .cmtf import perform_CMTF
+from tfac.cmtf import perform_CMTF
 
 PATH_HERE = dirname(dirname(abspath(__file__)))
 OPTIMAL_SCALING = 2 ** 7.0

tfac/figures/JamesGenerator.py

@@ -0,0 +1,28 @@
+import importlib
+from matplotlib.pyplot import close
+
+figureSuffixes = ["2", "4", "5", "6", "A1", "A2", "S1", "S2", "S3",
+                  "S4", "S5"]
+
+figModules = []
+for _ in figureSuffixes:
+    figModules.append(importlib.import_module("figure"+_))
+    # print(figModules)
+
+
+for figure in figModules[2:4]:
+    if hasattr(figure, "makeFigure"):
+        print(f"Calling: {figure.__name__}.makefigure()")
+        print(figure)
+        fig = figure.makeFigure()
+        fig.savefig(f"./JamesIter{figure.__name__}.png")
+
+
+"""
+There's a bug with figure6.makeFigure(). It fails to make the figure if
+the loop had just previously generated figure5. Somehow, there's a leftover
+'<' somewhere in the data stream which causes predict.py to fail
+
+Figure S5 takes a long time to make with no print statements indicating
+progress. 
+"""

tfac/figures/figure2.py

@@ -4,10 +4,11 @@
 import numpy as np
 import pandas as pd
 
-from .common import getSetup
-from ..dataImport import form_tensor, get_factors
-from ..predict import run_model
-from ..cmtf import calcR2X, PCArand
+
+from tfac.figures.common import getSetup
+from tfac.dataImport import form_tensor, get_factors
+from tfac.predict import run_model
+from tfac.cmtf import calcR2X, PCArand
 
 
 def get_r2x_results():

tfac/figures/figure4.py

@@ -11,10 +11,10 @@
 from sklearn.preprocessing import scale
 from sklearn.utils import resample
 
-from .common import getSetup
-from ..dataImport import form_tensor, import_cytokines, get_factors, \
+from tfac.figures.common import getSetup
+from tfac.dataImport import form_tensor, import_cytokines, get_factors, \
     reorder_table
-from ..predict import run_model, predict_regression
+from tfac.predict import run_model, predict_regression
 
 N_BOOTSTRAP = 30
 PATH_HERE = dirname(dirname(abspath(__file__)))

tfac/figures/figure5.py

@@ -9,10 +9,10 @@
 import pandas as pd
 from sklearn.metrics import roc_curve
 
-from .common import getSetup
-from ..dataImport import import_validation_patient_metadata, get_factors, \
+from tfac.figures.common import getSetup
+from tfac.dataImport import import_validation_patient_metadata, get_factors, \
     import_cytokines, import_rna
-from ..predict import get_accuracy, predict_known
+from tfac.predict import get_accuracy, predict_known
 
 COLOR_CYCLE = matplotlib.rcParams['axes.prop_cycle'].by_key()['color']
 PATH_HERE = dirname(dirname(abspath(__file__)))

tfac/figures/figure6.py

@@ -3,9 +3,9 @@
 import numpy as np
 import pandas as pd
 
-from .common import getSetup
-from ..dataImport import get_factors
-from ..predict import predict_known
+from tfac.figures.common import getSetup
+from tfac.dataImport import get_factors
+from tfac.predict import predict_known
 
 
 def makeFigure():

tfac/figures/figureA1.py

@@ -4,8 +4,8 @@
 import pandas as pd
 import numpy as np
 
-from .common import subplotLabel, getSetup
-from ..impute import evaluate_missing
+from tfac.figures.common import subplotLabel, getSetup
+from tfac.impute import evaluate_missing
 
 
 def makeFigure():

tfac/figures/figureA2.py

@@ -8,8 +8,8 @@
 import pandas as pd
 import seaborn as sns
 
-from .common import getSetup
-from ..dataImport import get_factors, import_rna, reorder_table
+from tfac.figures.common import getSetup
+from tfac.dataImport import get_factors, import_rna, reorder_table
 
 plt.rcParams["svg.fonttype"] = "none"
 

tfac/figures/figureS1.py

@@ -7,8 +7,8 @@
 from scipy.stats import pearsonr
 import seaborn as sns
 
-from .common import getSetup
-from ..dataImport import form_tensor, import_cytokines
+from tfac.figures.common import getSetup
+from tfac.dataImport import form_tensor, import_cytokines
 
 
 def adjust_box_widths(g, fac):

tfac/figures/figureS2.py

@@ -5,8 +5,8 @@
 from matplotlib.patches import Patch
 import numpy as np
 
-from .common import getSetup
-from ..dataImport import import_cytokines
+from tfac.figures.common import getSetup
+from tfac.dataImport import import_cytokines
 
 COLOR_CYCLE = matplotlib.rcParams['axes.prop_cycle'].by_key()['color']
 

tfac/figures/figureS3.py

@@ -8,10 +8,10 @@
 import pandas as pd
 from sklearn.metrics import r2_score, roc_curve
 
-from .common import getSetup
-from ..cmtf import OPTIMAL_RANK
-from ..dataImport import import_validation_patient_metadata, get_factors
-from ..predict import get_accuracy, predict_known, predict_validation, \
+from tfac.figures.common import getSetup
+from tfac.cmtf import OPTIMAL_RANK
+from tfac.dataImport import import_validation_patient_metadata, get_factors
+from tfac.predict import get_accuracy, predict_known, predict_validation, \
     predict_regression
 
 COLOR_CYCLE = matplotlib.rcParams['axes.prop_cycle'].by_key()['color'][2:]

tfac/figures/figureS4.py

@@ -6,8 +6,8 @@
 from scipy.stats import pearsonr
 import seaborn as sns
 
-from .common import getSetup
-from ..dataImport import import_cytokines
+from tfac.figures.common import getSetup
+from tfac.dataImport import import_cytokines
 
 
 def correlate_cytokines():

tfac/figures/figureS5.py

@@ -6,9 +6,9 @@
 from os.path import abspath, dirname
 import pandas as pd
 
-from .common import getSetup
-from ..dataImport import import_validation_patient_metadata, get_factors, import_cytokines, import_rna
-from ..predict import get_accuracy, predict_known, predict_validation, predict_regression
+from tfac.figures.common import getSetup
+from tfac.dataImport import import_validation_patient_metadata, get_factors, import_cytokines, import_rna
+from tfac.predict import get_accuracy, predict_known, predict_validation, predict_regression
 
 COLOR_CYCLE = matplotlib.rcParams['axes.prop_cycle'].by_key()['color'][2:]
 PATH_HERE = dirname(dirname(abspath(__file__)))

tfac/impute.py

@@ -2,8 +2,8 @@
 
 import numpy as np
 from statsmodels.multivariate.pca import PCA
-from .dataImport import form_tensor
-from .cmtf import perform_CMTF, calcR2X
+from tfac.dataImport import form_tensor
+from tfac.cmtf import perform_CMTF, calcR2X
 
 
 def flatten_to_mat(tensor, matrix=None):

tfac/predict.py

@@ -9,7 +9,7 @@
     RepeatedStratifiedKFold, StratifiedKFold
 from sklearn.svm import SVC
 
-from .dataImport import import_validation_patient_metadata
+from tfac.dataImport import import_validation_patient_metadata
 
 warnings.filterwarnings('ignore', category=UserWarning)
 
@@ -95,6 +95,7 @@ def predict_known(data, labels, method='predict', svc=False):
     if svc:
         _, model = run_svc(data, labels)
     else:
+
         _, model = run_model(data, labels)
 
     if isinstance(data, pd.Series):

tfac/tests/james_test.py

@@ -0,0 +1,72 @@
+import numpy as np
+
+def mode_n_product(tensor:np.ndarray, matrix:np.ndarray, mode:int):
+    """
+    Perform mode-n product of a tensor with a matrix.
+    
+    Parameters:
+    tensor (ndarray): Input tensor of shape (I1, I2, ..., In, ..., IN)
+    matrix (ndarray): Matrix to multiply with, of shape (J, In)
+    mode (int): Mode along which to multiply (1-based index)
+    
+    Returns:
+    ndarray: Resulting tensor after mode-n multiplication
+    """
+    # Convert to 0-based index for internal operations
+    mode -= 1
+
+    # Check if dimensions are compatible for multiplication
+    if tensor.shape[mode] != matrix.shape[1]:
+        raise ValueError(f"Cannot multiply: tensor's mode-{mode+1} dimension ({tensor.shape[mode]}) "
+                         f"does not match matrix's second dimension ({matrix.shape[1]}).")
+
+    # Move the mode-n dimension to the first dimension
+    tensor = np.moveaxis(tensor, mode, 0)
+
+    # Unfold the tensor into a matrix
+    unfolded_tensor = tensor.reshape(tensor.shape[0], -1)
+
+    # Perform matrix multiplication
+    result:np.ndarray = np.dot(matrix, unfolded_tensor)
+
+    # Fold the resulting matrix back into a tensor
+    # new_shape = list(tensor.shape)
+    # new_shape[0] = matrix.shape[0]
+    # new_shape = tuple(new_shape)
+    # result_tensor = result.reshape(new_shape)
+    result_tensor = result.reshape((matrix.shape[0], tensor.shape[1], tensor.shape[2]))
+
+    # Move the dimensions back to the original order
+    result_tensor = np.moveaxis(result_tensor, 0, mode)
+    return result_tensor
+
+# Example tensor (2 x 3 x 4)
+tensor = np.random.rand(2, 3, 4)
+print("Original Tensor:")
+print(tensor)
+
+# Example matrix (5 x 3) to multiply along mode-2 (1-based index)
+matrix = np.random.rand(5, 3)
+print("\nMatrix:")
+print(matrix)
+
+try:
+    # Perform mode-2 (1-based index) multiplication
+    result_tensor = mode_n_product(tensor, matrix, 2)
+    print("\nResulting Tensor after mode-2 multiplication:")
+    print(result_tensor)
+except ValueError as e:
+    print(f"\nError: {e}")
+
+# Example of incompatible matrix (5 x 7)
+incompatible_matrix = np.random.rand(5, 7)
+print("\nIncompatible Matrix:")
+print(incompatible_matrix)
+
+try:
+    # Attempt to perform mode-2 multiplication with an incompatible matrix
+    result_tensor = mode_n_product(tensor, incompatible_matrix, 2)
+    print("\nResulting Tensor after mode-2 multiplication with incompatible matrix:")
+    print(result_tensor)
+except ValueError as e:
+    print(f"\nError: {e}")

tfac/tests/test_import.py

@@ -4,7 +4,7 @@
 import pytest
 import numpy as np
 import pandas as pd
-from ..dataImport import import_patient_metadata, form_tensor, import_rna
+from tfac.dataImport import import_patient_metadata, form_tensor, import_rna
 
 
 @pytest.mark.parametrize("call", [import_patient_metadata, import_rna])