Update monai spleen example and monai_nvflare [skip ci] (#1083)

* update error message

* update setup; add example summary stats

* update monai section of root readme

* fix typo

* increase rc version

* increase rc version

Author: holgerroth

README.md

@@ -60,9 +60,9 @@ pip install nvflare
   The results, for all features of all datasets at all sites as well as global aggregates, can be visualized via the visualization utility in the notebook.  
 
 * [MONAI Integration](<https://github.com/NVIDIA/NVFlare/tree/main/integration/monai/README.md>)
-  In 2.2 release, we provided two implementations by leveraging Monai "bundle"
-  * Monai [ClientAlgo](https://docs.monai.io/en/latest/fl.html#monai.fl.client.ClientAlgo) Integration -- enable running MONAI bundles directly in a federated setting using NVFLARE
-  * Monai [Statistics](https://docs.monai.io/en/latest/fl.html#monai.fl.client.ClientAlgoStats) Integration -- through NVFLARE Federated Statistics we can generate, compare and visualize all client's data statistics generated from Monai summary statistics
+  In 2.2 release, we provided two implementations by leveraging MONAI [Bundle](https://docs.monai.io/en/latest/bundle_intro.html).
+  * MONAI [ClientAlgo](https://docs.monai.io/en/latest/fl.html#monai.fl.client.ClientAlgo) Integration -- enable running MONAI bundles directly in a federated setting using NVFLARE
+  * MONAI [ClientAlgoStats](https://docs.monai.io/en/latest/fl.html#monai.fl.client.ClientAlgoStats) Integration -- through NVFLARE Federated Statistics we can generate, compare and visualize all clients' data statistics generated from MONAI summary statistics
 
 * Tools and Production Support
   * [Improved POC command](https://nvflare.readthedocs.io/en/main/user_guide/poc_command.html) 
@@ -74,7 +74,7 @@ pip install nvflare
 ### Migrations tips 
 
    To migrate from releases prior to 2.2.1, here are few notes that might help
-   [migrate to 2.2.1](docs/release_notes/2.2.1/migration_notes.md)
+   [migrate to 2.2.1](docs/release_notes/2.2.1/migration_notes.md).
 
 
 ## Third party license

integration/monai/examples/spleen_ct_segmentation/README.md

@@ -95,7 +95,9 @@ To avoid caching the data for training (which will not be executed during the st
 ```
 nvflare simulator job_stats --workspace /tmp/nvflare/sim_spleen_ct_seg --threads 2 --n_clients 2 --gpu 0,1
 ```
-The results will be placed in the workspace directory under `simulate_job/statistics/image_statistics.json`.
+The histogram results will be placed in the workspace directory under `simulate_job/statistics/image_statistics.json`.
+Additional summary statistics, like average image shapes and spacings, computed by MONAI are placed under `simulate_job/app_server/`.
+See [stats_demo/simulate_job/app_server](./stats_demo/simulate_job/app_server) for examples.
 
 For an end-to-end demo, including visualization of the gathered statistics, start a Jupyter Lab with `stats_demo`
 ```

integration/monai/examples/spleen_ct_segmentation/stats_demo/histogram_main.png

@@ -1,8 +1,8 @@
-tensorboard
 tqdm
 nibabel
 fire
 pytorch-ignite>=0.4.10
 monai>=1.0.1rc3
-nvflare>=2.2.1rc8
-monai_nvflare>=0.2.0
+nvflare>=2.2.1rc9
+monai_nvflare>=0.2.1rc1
+tensorboard

integration/monai/examples/spleen_ct_segmentation/stats_demo/histogram_subplot.png

@@ -103,7 +103,8 @@ def pre_run(
                     raise ValueError(
                         f"There are more returned histograms ({n_hists}) for dataset "
                         f"{dataset_name} than provided feature names. "
-                        f"Please use '*' to define the histogram bins and range for all features."
+                        f"Please use '*' to define the histogram bins and range for all features "
+                        f"or provide histograms bins and range for each feature."
                     )
                 if fn == "*":
                     fn = "Intensity"
@@ -117,7 +118,7 @@ def pre_run(
             if len(self.feature_names) != n_hists:
                 raise ValueError(
                     f"Given length of feature names {self.feature_names} ({len(self.feature_names)}) "
-                    f"do not match returned histograms ({n_hists})!"
+                    f"do not match returned histograms ({n_hists}) for dataset {dataset_name}!"
                 )
             for _hist_fn, _histo in zip(self.feature_names, hist_list):
                 self.histograms[dataset_name][_hist_fn] = _histo

integration/monai/examples/spleen_ct_segmentation/stats_demo/simulate_job/app_server/site-1_data_stats.yaml

@@ -55,5 +55,5 @@
     long_description=long_description,
     long_description_content_type="text/markdown",
     python_requires=">=3.7,<3.9",
-    install_requires=["monai==1.0.1rc3", "nvflare>=2.2.1rc8"],
+    install_requires=["monai>=1.0.1rc4", "nvflare>=2.2.1rc9"],
 )