Merge pull request #240 from EducationalTestingService/update-stats-m…

…odels-and-seaborn

Update versions for `statsmodels` and `seaborn`

conda_requirements.txt

@@ -9,9 +9,9 @@ numpy=1.13.3
 pandas=0.22.0
 scikit-learn=0.19.1
 scipy=1.0.0
-seaborn=0.8.1
+seaborn=0.9.0
 skll=1.5.2
-statsmodels=0.8.0
+statsmodels=0.9.0
 sphinx
 sphinx_rtd_theme
 zeromq

environment.yml

@@ -14,9 +14,9 @@ dependencies:
 - pandas=0.22.0
 - scikit-learn=0.19.1
 - scipy=1.0.0
-- seaborn=0.8.1
+- seaborn=0.9.0
 - skll=1.5.2
-- statsmodels=0.8.0
+- statsmodels=0.9.0
 - parameterized
 - zeromq
 - sphinx

requirements.txt

@@ -8,9 +8,9 @@ numpy==1.13.3
 pandas==0.22.0
 scikit-learn==0.19.1
 scipy==1.0.0
-seaborn==0.8.1
+seaborn==0.9.0
 skll==1.5.2
-statsmodels==0.8.0
+statsmodels==0.9.0
 openpyxl
 xlrd
 xlwt

rsmtool/notebooks/dff_by_group.ipynb

@@ -36,9 +36,10 @@
     "    group_values = sorted(df_melted[group].unique())\n",
     "    colors = sns.color_palette(\"Greys\", len(group_values))\n",
     "    with sns.axes_style('whitegrid'), sns.plotting_context('notebook', font_scale=1.2):\n",
-    "        p = sns.factorplot(x='sc1', y='value', hue=group, hue_order = group_values,\n",
-    "                           col='feature', col_wrap=3, size=3, scale=0.6, palette=colors,\n",
-    "                           sharey=False, sharex=False, legend=False, data=df_melted)\n",
+    "        p = sns.catplot(x='sc1', y='value', hue=group, hue_order = group_values,\n",
+    "                        col='feature', col_wrap=3, height=3, scale=0.6, palette=colors,\n",
+    "                        sharey=False, sharex=False, legend=False, kind=\"point\",\n",
+    "                        data=df_melted)\n",
     "\n",
     "        for i, axis in enumerate(p.axes):\n",
     "            axis.set_xlabel('score')\n",

rsmtool/notebooks/evaluation.ipynb

@@ -181,9 +181,9 @@
     "\n",
     "    # make a barplot without a legend since we will \n",
     "    # add one manually later\n",
-    "    p = sns.factorplot(\"score\", \"value\", \"variable\", kind=\"bar\",\n",
-    "                       palette=colors, data=df_scoredist_melted, \n",
-    "                       size=3, aspect=2, legend=False)\n",
+    "    p = sns.catplot(\"score\", \"value\", \"variable\", kind=\"bar\",\n",
+    "                    palette=colors, data=df_scoredist_melted, \n",
+    "                    height=3, aspect=2, legend=False)\n",
     "    p.set_axis_labels('score', '% of responses')\n",
     "    \n",
     "    # add a legend with the right colors\n",
@@ -233,7 +233,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.6.4"
+   "version": "3.6.3"
   }
  },
  "nbformat": 4,

rsmtool/notebooks/feature_descriptives.ipynb

@@ -68,9 +68,9 @@
     "with sns.axes_style('whitegrid'):\n",
     "    # create a barplot without a legend since we will manually\n",
     "    # add one later\n",
-    "    p = sns.factorplot(\"feature\", \"value\", \"variable\", kind=\"bar\", \n",
-    "                       palette=colors, data=df_outliers, size=3, \n",
-    "                       aspect=aspect, legend=False)\n",
+    "    p = sns.catplot(\"feature\", \"value\", \"variable\", kind=\"bar\", \n",
+    "                    palette=colors, data=df_outliers, height=3, \n",
+    "                    aspect=aspect, legend=False)\n",
     "    p.set_axis_labels('', '% cases truncated to mean +/- 4*sd')\n",
     "    p.set_xticklabels(rotation=90)\n",
     "    p.set(ylim=limits)\n",

rsmtool/notebooks/preprocessed_features.ipynb

@@ -47,7 +47,7 @@
     "    warnings.filterwarnings('ignore', category=np.VisibleDeprecationWarning)\n",
     "    with sns.axes_style('white'):\n",
     "        g = sns.FacetGrid(col='variable', data=df_train_preproc_melted, col_wrap=3, \n",
-    "                          col_order=column_order, sharex=False, sharey=False, size=2, \n",
+    "                          col_order=column_order, sharex=False, sharey=False, height=2, \n",
     "                          aspect=1)\n",
     "        g.map(sns.distplot, \"value\", color=\"grey\", kde=False)\n",
     "        for ax, cname in zip(g.axes, g.col_names):\n",
@@ -181,9 +181,9 @@
     "\n",
     "    # generate a bar plot but without the legend since we will\n",
     "    # manually add one later\n",
-    "    p = sns.factorplot(\"feature\", \"value\", \"variable\", kind=\"bar\",\n",
-    "                       palette=colors, data=df_mpcor, size=3, \n",
-    "                       aspect=aspect, legend=False)\n",
+    "    p = sns.catplot(\"feature\", \"value\", \"variable\", kind=\"bar\",\n",
+    "                    palette=colors, data=df_mpcor, height=3, \n",
+    "                    aspect=aspect, legend=False)\n",
     "    p.set_axis_labels('', 'Correlation with score')\n",
     "    p.set_xticklabels(rotation=90)\n",
     "    p.set(ylim=limits)\n",
@@ -252,9 +252,9 @@
     "        \n",
     "        # create a barplot but without the legend since\n",
     "        # we will manually add one later\n",
-    "        p = sns.factorplot(\"feature\", \"value\", \"variable\", kind=\"bar\",\n",
-    "                           palette=colors, data=df_mpcor, size=3, \n",
-    "                           aspect=aspect, legend=False)\n",
+    "        p = sns.catplot(\"feature\", \"value\", \"variable\", kind=\"bar\",\n",
+    "                        palette=colors, data=df_mpcor, height=3, \n",
+    "                        aspect=aspect, legend=False)\n",
     "        p.set_axis_labels('', 'Correlation with length')\n",
     "        p.set_xticklabels(rotation=90)\n",
     "        p.set(ylim=limits)\n",