mihnathul commited on
Commit
71a631c
·
verified ·
1 Parent(s): 5b9539c

docs: prominent download instructions

Browse files
Files changed (1) hide show
  1. README.md +22 -15
README.md CHANGED
@@ -72,9 +72,28 @@ negative = discharge.
72
  Join keys: `cells.cell_id = tests.cell_id`, `tests.test_id = timeseries.test_id`,
73
  `(tests.test_id, cycle_number) = cycle_summary.(test_id, cycle_number)`.
74
 
75
- ## Quick start
76
 
77
- ### DuckDB — read everything in place, no download
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
78
 
79
  ```sql
80
  INSTALL httpfs; LOAD httpfs;
@@ -90,23 +109,20 @@ GROUP BY 1,2,3,4,5
90
  ORDER BY t.test_id;
91
  ```
92
 
93
- ### pandas — predicate-pushdown read of one test
94
 
95
  ```python
96
  import pandas as pd
97
-
98
  df = pd.read_parquet(
99
  "https://huggingface.co/datasets/mihnathul/celljar/resolve/main/timeseries.parquet",
100
  filters=[("test_id", "==", "ORNL_LEAF_2013_HPPC_25C")],
101
  )
102
- print(df.head())
103
  ```
104
 
105
  ### `datasets` library — streaming
106
 
107
  ```python
108
  from datasets import load_dataset
109
-
110
  ds = load_dataset(
111
  "parquet",
112
  data_files="https://huggingface.co/datasets/mihnathul/celljar/resolve/main/timeseries.parquet",
@@ -117,15 +133,6 @@ for row in ds.take(5):
117
  print(row)
118
  ```
119
 
120
- ### `huggingface_hub` — pull the whole bundle locally
121
-
122
- ```python
123
- from huggingface_hub import snapshot_download
124
-
125
- local = snapshot_download(repo_id="mihnathul/celljar", repo_type="dataset")
126
- print(local) # contains cells/, tests/, timeseries.parquet
127
- ```
128
-
129
  ## License & citation
130
 
131
  The science here belongs to the original authors; celljar simply puts their
 
72
  Join keys: `cells.cell_id = tests.cell_id`, `tests.test_id = timeseries.test_id`,
73
  `(tests.test_id, cycle_number) = cycle_summary.(test_id, cycle_number)`.
74
 
75
+ ## Download the whole bundle
76
 
77
+ ```bash
78
+ # CLI — pulls everything (cells/*.json, tests/*.json, timeseries.parquet, cycle_summary.parquet)
79
+ pip install huggingface_hub
80
+ huggingface-cli download mihnathul/celljar --repo-type dataset --local-dir ./celljar-bundle
81
+
82
+ # Pin a tagged release for reproducibility
83
+ huggingface-cli download mihnathul/celljar --repo-type dataset --revision v0.2.0 --local-dir ./celljar-bundle
84
+ ```
85
+
86
+ Or in Python:
87
+
88
+ ```python
89
+ from huggingface_hub import snapshot_download
90
+ local = snapshot_download(repo_id="mihnathul/celljar", repo_type="dataset", revision="v0.2.0")
91
+ print(local) # local path containing cells/, tests/, timeseries.parquet, cycle_summary.parquet
92
+ ```
93
+
94
+ ## Query in place — no download needed
95
+
96
+ ### DuckDB — full SQL across all entities over HTTPS
97
 
98
  ```sql
99
  INSTALL httpfs; LOAD httpfs;
 
109
  ORDER BY t.test_id;
110
  ```
111
 
112
+ ### pandas / Polars — predicate-pushdown read of one test
113
 
114
  ```python
115
  import pandas as pd
 
116
  df = pd.read_parquet(
117
  "https://huggingface.co/datasets/mihnathul/celljar/resolve/main/timeseries.parquet",
118
  filters=[("test_id", "==", "ORNL_LEAF_2013_HPPC_25C")],
119
  )
 
120
  ```
121
 
122
  ### `datasets` library — streaming
123
 
124
  ```python
125
  from datasets import load_dataset
 
126
  ds = load_dataset(
127
  "parquet",
128
  data_files="https://huggingface.co/datasets/mihnathul/celljar/resolve/main/timeseries.parquet",
 
133
  print(row)
134
  ```
135
 
 
 
 
 
 
 
 
 
 
136
  ## License & citation
137
 
138
  The science here belongs to the original authors; celljar simply puts their