dump autosk

Ciaran1981 · Jul 20, 2023 · 01fdd76 · 01fdd76
1 parent 9aafc1b
commit 01fdd76
Show file tree

Hide file tree

Showing 7 changed files with 452 additions and 731 deletions.
diff --git a/geolearn_env.yml b/geolearn_env.yml
@@ -1,68 +1,54 @@
-name: geospatial_learn_
+name: geolearn
 
 channels:
   - pytorch
   - conda-forge
-  - anaconda
+
 
 dependencies:
-  - python=3.7 #tpot 
-  - pip
-  - numpy
+  - python=3 
+  - scikit-image
+  - scikit-learn
   - pytorch-gpu
   - torchvision
   - segmentation-models-pytorch
   - jupyter
   - matplotlib
   - opencv
   - gdal
-  - scikit-image
-  - scikit-learn
+  - lightgbm
   - git
   - pandas
   - albumentations
-  - imageio
   - ipywidgets
   - nb_conda
-  - joblib
-  - glob2
-  - pillow
-  - tqdm
-  - joblib
-  - lxml
   - pandas
-  - glob2
-  - more-itertools
-  - psutil
   - shapely
-  - tpot
   - xgboost
   - pysal
   - pyshp
   - pytables
-  - sqlalchemy
-  - urbanaccess
-  - pyntcloud
+  - mahotas
+#  - sqlalchemy ?
+#  - urbanaccess ?
   - napari
   - scikit-plot
   - dask-ml
-  - auto-sklearn
+#  - auto-sklearn requires too old sklearn
   - ipywidgets
   - nb_conda
   - owslib
   - optuna
-
+  - plotly
 
   - pip:
-    - xmltodict
     - morphsnakes
+    - xmltodict
     - simpledbf
     - pyfftw
     - phasepack
-    - mahotas
-    - plyfile
+
     - SimpleCRF
-    - skorch
     - git+https://github.com/Ciaran1981/geospatial-learn#egg=geospatial-learn
 #py
 

diff --git a/geolearn_env_new.yml b/geolearn_env_new.yml
diff --git a/geospatial_learn/handyplots.py b/geospatial_learn/handyplots.py
@@ -19,7 +19,7 @@
 import geopandas as gpd
 from sklearn.metrics import accuracy_score, classification_report, confusion_matrix,jaccard_score, f1_score
 from sklearn.metrics import precision_recall_fscore_support as prf
-from sklearn.metrics import plot_confusion_matrix
+#from sklearn.metrics import plot_confusion_matrix
 import scikitplot as skplt
 import seaborn as sns
 import pandas as pd
@@ -123,77 +123,78 @@ def _plt_heatmap(values, xlabel, ylabel, xticklabels, yticklabels,
         ax.text(x, y, fmt % value, color=c, ha="center", va="center")
     return img
 
-def plt_confmat(X_test, y_test, model, class_names=None, cmap=plt.cm.Blues, 
-                fmt="%d", save=None):
+#TODO rewrite
+# def plt_confmat(X_test, y_test, model, class_names=None, cmap=plt.cm.Blues, 
+#                 fmt="%d", save=None):
 
 
-    """
-    Plot a confusion matrix
+#     """
+#     Plot a confusion matrix
 
-    Parameters
-    -------------------
+#     Parameters
+#     -------------------
 
-    trueVals : nparray
-        array of reference/training/validation values
+#     trueVals : nparray
+#         array of reference/training/validation values
 
-    predVals : nparray
-        the predicted values
+#     predVals : nparray
+#         the predicted values
 
-    cmap : matplot lib object (optional)
-        eg plt.cm.gray
+#     cmap : matplot lib object (optional)
+#         eg plt.cm.gray
 
-    Returns:
+#     Returns:
 
-    The confusion matrix and a plot
-    """
-    #labels = np.unique(trueVals)
-    # the above heatmap function is used to create the plot
+#     The confusion matrix and a plot
+#     """
+#     #labels = np.unique(trueVals)
+#     # the above heatmap function is used to create the plot
 
 
-    #skplt.metrics.plot_confusion_matrix(trueVals, predVals, normalize=True)
+#     #skplt.metrics.plot_confusion_matrix(trueVals, predVals, normalize=True)
 
 
 
-    #conf = confusion_matrix(trueVals, predVals)
+#     #conf = confusion_matrix(trueVals, predVals)
 
-#    titles_options = [("Confusion matrix, without normalization", None),
-#                  ("Normalized confusion matrix", 'true')]
+# #    titles_options = [("Confusion matrix, without normalization", None),
+# #                  ("Normalized confusion matrix", 'true')]
 
-    confs = []
-    disps = []
-    #first is gonna be several digits, second a decimal of max 2
+#     confs = []
+#     disps = []
+#     #first is gonna be several digits, second a decimal of max 2
 
-    font_sizes = [{'font.size': 5}, {'font.size': 5}] 
-    # one is a count the other a decimal
-    vformats = ['.0f', '.2f']
+#     font_sizes = [{'font.size': 5}, {'font.size': 5}] 
+#     # one is a count the other a decimal
+#     vformats = ['.0f', '.2f']
 
-    titles = ["Confusion matrix, without normalization", 
-               "Normalized confusion matrix"]
-    normalise = [None, 'true']
+#     titles = ["Confusion matrix, without normalization", 
+#                "Normalized confusion matrix"]
+#     normalise = [None, 'true']
 
-    #fig, ax = plt.subplots(figsize=(10, 10))
+#     #fig, ax = plt.subplots(figsize=(10, 10))
 
-    for t, n,f,v in zip(titles, normalise, font_sizes, vformats):
-        disp = plot_confusion_matrix(model, X_test, y_test,
-                                     display_labels=class_names,
-                                     cmap=cmap,
-                                     normalize=n,
-                                     xticks_rotation='vertical',
-                                     values_format=v)
-        plt.rcParams.update(f)
-        disp.ax_.set_title(t)
-        disps.append(disp)
-        confs.append(disp.confusion_matrix)
+#     for t, n,f,v in zip(titles, normalise, font_sizes, vformats):
+#         disp = plot_confusion_matrix(model, X_test, y_test,
+#                                      display_labels=class_names,
+#                                      cmap=cmap,
+#                                      normalize=n,
+#                                      xticks_rotation='vertical',
+#                                      values_format=v)
+#         plt.rcParams.update(f)
+#         disp.ax_.set_title(t)
+#         disps.append(disp)
+#         confs.append(disp.confusion_matrix)
 
-    plt.show()
+#     plt.show()
 
-    types = ['cnts','nrm']
-    for t, d in zip(types, disps):
-        # must set a size
-#        d.figure_(figsize=(18, 9))
-        d.figure_.savefig(save[:-3]+t+'.png', bbox_inches='tight')
+#     types = ['cnts','nrm']
+#     for t, d in zip(types, disps):
+#         # must set a size
+# #        d.figure_(figsize=(18, 9))
+#         d.figure_.savefig(save[:-3]+t+'.png', bbox_inches='tight')
 
-    return confs
+#     return confs