Merge pull request #19 from individual-brain-charting/new_versions

Add new versions of IBC database

Author: man-shu

src/ibc_api/data/preprocessed_v2.csv

@@ -0,0 +1 @@
+,subject,session,desc,hemi,task,direction,run,space,suffix,datatype,extension,contrast,megabytes,dataset,path

src/ibc_api/data/raw_v4.csv

@@ -1,31 +1,41 @@
 """Script to create the database from the raw, preprocessed, volume maps and surface maps
- data on EBRAINS."""
+data on EBRAINS."""
 
 # import libraries
 import pandas as pd
 import os
 import ibc_api.utils as ibc
+from ibc_api.metadata import fetch_metadata, _find_latest_version
 from bids.layout import parse_file_entities
 
 datasets = ["raw", "preprocessed", "volume_maps", "surface_maps"]
 
-ibc.authenticate()
+ibc._authenticate()
 for dataset in datasets:
-    for version in range(1, 4):
+    for version in range(1, 6):
         # Get EBRAINS metadata about the dataset
         try:
             ebrains_data = ibc._connect_ebrains(dataset, version=version)
         except (ValueError, IndexError) as error:
+            print(error)
             print(f"skipping dataset {dataset}, version {version}")
             continue
+        try:
+            root_dir = ebrains_data.prefix.strip("/")
+        except AttributeError:
+            root_dir = ""
         # Get the file names and other info as dataframes
         ebrains_df = pd.DataFrame(ebrains_data.__dict__["_files"])
         filenames = ebrains_df["name"].tolist()
         # parse filenames using pybids to get all the entities
         bids_entities = []
         for file in filenames:
             bids_entity = parse_file_entities(
-                file, include_unmatched=True, config="ibc_conifg.json"
+                file,
+                include_unmatched=True,
+                config=os.path.join(
+                    os.path.dirname(__file__), "ibc_config.json"
+                ),
             )
             bids_entities.append(bids_entity)
         # convert the list of dictionaries with bids entities to a dataframe
@@ -36,25 +46,39 @@
         bids_df["megabytes"] = ebrains_df["bytes"].astype(int).div(1024**2)
         # add a column with the dataset name
         bids_df["dataset"] = [dataset] * len(bids_df)
-        root_dir = ebrains_df["name"].str.split("/").str[0]
-        # add a column with the file path without the root directory
-        path = ebrains_df["name"].str.split("/").str[1:].str.join("/")
+        infer_root_dir = ebrains_df["name"].str.split("/").str[0].unique()[0]
+
+        if infer_root_dir == root_dir:
+            # add a column with the file path without the root directory
+            path = ebrains_df["name"].str.split("/").str[1:].str.join("/")
+        else:
+            # add a column with the file path
+            path = ebrains_df["name"]
+
+        breakpoint()
+
+        root_dir_series = ebrains_df["name"].str.split("/").str[0]
         bids_df["path"] = path
         # separate surface maps and volume maps in different csv files
         if dataset == "surface_maps":
-            mask = (root_dir == "resulting_smooth_maps_surface") & bids_df[
-                "extension"
-            ].isin([".gii", ".json"])
+            mask = (
+                root_dir_series == "resulting_smooth_maps_surface"
+            ) & bids_df["extension"].isin([".gii", ".json"])
             bids_df = bids_df[mask]
         # there are some files with .gii extension in the volume maps folder
         # filtering them out
         elif dataset == "volume_maps":
-            mask = (root_dir == "resulting_smooth_maps") & bids_df[
+            mask = (root_dir_series == "resulting_smooth_maps") & bids_df[
                 "extension"
             ].isin([".nii.gz", ".json"])
             bids_df = bids_df[mask]
         bids_df = bids_df.reset_index(drop=True)
         # create a csv file with the bids entities
-        csv_file = os.path.join("..", "data", f"{dataset}_v{version}.csv")
+        csv_file = os.path.join(
+            os.path.dirname(__file__),
+            "..",
+            "data",
+            f"{dataset}_v{version}.csv",
+        )
         bids_df.to_csv(csv_file)
         print(f"{csv_file} created!")

src/ibc_api/data/raw_v5.csv

@@ -50,7 +50,7 @@ def _authenticate(token_dir=TOKEN_ROOT):
         with open(token_file, "r") as f:
             token = f.read()
             # set the token
-            siibra.set_ebrains_token(token)
+            EbrainsRequest.set_token(token)
     else:
         EbrainsRequest.fetch_token()
         token = EbrainsRequest._KG_API_TOKEN