Skip to content

Commit

Permalink
[pre-commit.ci] auto fixes from pre-commit.com hooks
Browse files Browse the repository at this point in the history 
for more information, see https://pre-commit.ci
  • Loading branch information
@pre-commit-ci
pre-commit-ci[bot] committed Sep 2, 2024
1 parent 2e39418 commit 1e0bcfc
Show file tree
Hide file tree
Showing 2 changed files with 130 additions and 55 deletions.
78 changes: 54 additions & 24 deletions example/fruits.ipynb
View file
Original file line number Diff line number Diff line change
Expand Up @@ -36,17 +36,46 @@
"metadata": {},
"outputs": [],
"source": [
"fruits = {'plants': ['fruit', 'fruit', 'fruit', 'fruit', 'vegetable', 'vegetable', 'vegetable', 'vegetable'],\n",
" 'plts': ['fr', 'fr', 'fr', 'fr', 've', 've', 've', 've'], \n",
" 'quantity': ['1 kg', '10 kg', '1 kg', '10 kg', '1 kg', '10 kg', '1 kg', '10 kg'],\n",
" 'product': ['apple', 'apple', 'orange', 'orange', 'peppers', 'peppers', 'carrot', 'carrot'],\n",
" 'price': [1, 10, 2, 20, 1.5, 15, 1.5, 20],\n",
" 'price level': ['low', 'low', 'high', 'high', 'low', 'low', 'high', 'high'],\n",
" 'group': ['fruit 1', 'fruit 10', 'fruit 1', 'veget', 'veget', 'veget', 'veget', 'veget'],\n",
" 'id': [1001, 1002, 1003, 1004, 1005, 1006, 1007, 1008],\n",
" 'supplier': [\"sup1\", \"sup1\", \"sup1\", \"sup2\", \"sup2\", \"sup2\", \"sup2\", \"sup1\"],\n",
" 'location': [\"fr\", \"gb\", \"es\", \"ch\", \"gb\", \"fr\", \"es\", \"ch\"],\n",
" 'valid': [\"ok\", \"ok\", \"ok\", \"ok\", \"ok\", \"ok\", \"ok\", \"ok\"]}"
"fruits = {\n",
" \"plants\": [\n",
" \"fruit\",\n",
" \"fruit\",\n",
" \"fruit\",\n",
" \"fruit\",\n",
" \"vegetable\",\n",
" \"vegetable\",\n",
" \"vegetable\",\n",
" \"vegetable\",\n",
" ],\n",
" \"plts\": [\"fr\", \"fr\", \"fr\", \"fr\", \"ve\", \"ve\", \"ve\", \"ve\"],\n",
" \"quantity\": [\"1 kg\", \"10 kg\", \"1 kg\", \"10 kg\", \"1 kg\", \"10 kg\", \"1 kg\", \"10 kg\"],\n",
" \"product\": [\n",
" \"apple\",\n",
" \"apple\",\n",
" \"orange\",\n",
" \"orange\",\n",
" \"peppers\",\n",
" \"peppers\",\n",
" \"carrot\",\n",
" \"carrot\",\n",
" ],\n",
" \"price\": [1, 10, 2, 20, 1.5, 15, 1.5, 20],\n",
" \"price level\": [\"low\", \"low\", \"high\", \"high\", \"low\", \"low\", \"high\", \"high\"],\n",
" \"group\": [\n",
" \"fruit 1\",\n",
" \"fruit 10\",\n",
" \"fruit 1\",\n",
" \"veget\",\n",
" \"veget\",\n",
" \"veget\",\n",
" \"veget\",\n",
" \"veget\",\n",
" ],\n",
" \"id\": [1001, 1002, 1003, 1004, 1005, 1006, 1007, 1008],\n",
" \"supplier\": [\"sup1\", \"sup1\", \"sup1\", \"sup2\", \"sup2\", \"sup2\", \"sup2\", \"sup1\"],\n",
" \"location\": [\"fr\", \"gb\", \"es\", \"ch\", \"gb\", \"fr\", \"es\", \"ch\"],\n",
" \"valid\": [\"ok\", \"ok\", \"ok\", \"ok\", \"ok\", \"ok\", \"ok\", \"ok\"],\n",
"}"
]
},
{
Expand All @@ -57,6 +86,7 @@
"outputs": [],
"source": [
"from tab_dataset import Sdataset\n",
"\n",
"dts = Sdataset.ntv(fruits)\n",
"adts = dts.analysis"
]
Expand Down Expand Up @@ -90,9 +120,9 @@
}
],
"source": [
"print(adts.get_relation('plants', 'plts').typecoupl)\n",
"print(adts.get_relation('plants', 'product').typecoupl)\n",
"print(adts.get_relation('quantity', 'product').typecoupl)\n"
"print(adts.get_relation(\"plants\", \"plts\").typecoupl)\n",
"print(adts.get_relation(\"plants\", \"product\").typecoupl)\n",
"print(adts.get_relation(\"quantity\", \"product\").typecoupl)"
]
},
{
Expand Down Expand Up @@ -123,11 +153,11 @@
}
],
"source": [
"print('minimum distance: ', adts.get_relation('plants', 'plts').distance)\n",
"print('maximum distance: ', adts.get_relation('id', 'valid').distance)\n",
"print('intermediate distance: ', adts.get_relation('plants', 'product').distance)\n",
"# The 'plants' - 'product' relationship will be 'coupled' if we change, for example, \n",
"#'fruit-orange' in 'citrus-orange' and 'carrot-vegetable' in 'carrot-root vegetable' (2 changes) "
"print(\"minimum distance: \", adts.get_relation(\"plants\", \"plts\").distance)\n",
"print(\"maximum distance: \", adts.get_relation(\"id\", \"valid\").distance)\n",
"print(\"intermediate distance: \", adts.get_relation(\"plants\", \"product\").distance)\n",
"# The 'plants' - 'product' relationship will be 'coupled' if we change, for example,\n",
"#'fruit-orange' in 'citrus-orange' and 'carrot-vegetable' in 'carrot-root vegetable' (2 changes)"
]
},
{
Expand Down Expand Up @@ -238,7 +268,7 @@
}
],
"source": [
"adts.partitions(mode='id')"
"adts.partitions(mode=\"id\")"
]
},
{
Expand Down Expand Up @@ -304,7 +334,7 @@
}
],
"source": [
"adts.field_partition(mode='id')"
"adts.field_partition(mode=\"id\")"
]
},
{
Expand Down Expand Up @@ -362,7 +392,7 @@
}
],
"source": [
"adts.field_partition(mode='id', partition=['product', 'quantity'])"
"adts.field_partition(mode=\"id\", partition=[\"product\", \"quantity\"])"
]
},
{
Expand Down Expand Up @@ -393,7 +423,7 @@
}
],
"source": [
"adts.relation_partition(partition=['product', 'quantity'])"
"adts.relation_partition(partition=[\"product\", \"quantity\"])"
]
},
{
Expand Down Expand Up @@ -438,7 +468,7 @@
"metadata": {},
"outputs": [],
"source": [
"print(dts.to_xarray(idxname=['product', 'quantity']))"
"print(dts.to_xarray(idxname=[\"product\", \"quantity\"]))"
]
}
],
Expand Down
107 changes: 76 additions & 31 deletions example/fruits_df.ipynb
View file
Original file line number Diff line number Diff line change
Expand Up @@ -36,17 +36,46 @@
"metadata": {},
"outputs": [],
"source": [
"fruits = {'plants': ['fruit', 'fruit', 'fruit', 'fruit', 'vegetable', 'vegetable', 'vegetable', 'vegetable'],\n",
" 'plts': ['fr', 'fr', 'fr', 'fr', 've', 've', 've', 've'], \n",
" 'quantity': ['1 kg', '10 kg', '1 kg', '10 kg', '1 kg', '10 kg', '1 kg', '10 kg'],\n",
" 'product': ['apple', 'apple', 'orange', 'orange', 'peppers', 'peppers', 'carrot', 'carrot'],\n",
" 'price': [1, 10, 2, 20, 1.5, 15, 1.5, 20],\n",
" 'price level': ['low', 'low', 'high', 'high', 'low', 'low', 'high', 'high'],\n",
" 'group': ['fruit 1', 'fruit 10', 'fruit 1', 'veget', 'veget', 'veget', 'veget', 'veget'],\n",
" 'id': [1001, 1002, 1003, 1004, 1005, 1006, 1007, 1008],\n",
" 'supplier': [\"sup1\", \"sup1\", \"sup1\", \"sup2\", \"sup2\", \"sup2\", \"sup2\", \"sup1\"],\n",
" 'location': [\"fr\", \"gb\", \"es\", \"ch\", \"gb\", \"fr\", \"es\", \"ch\"],\n",
" 'valid': [\"ok\", \"ok\", \"ok\", \"ok\", \"ok\", \"ok\", \"ok\", \"ok\"]}"
"fruits = {\n",
" \"plants\": [\n",
" \"fruit\",\n",
" \"fruit\",\n",
" \"fruit\",\n",
" \"fruit\",\n",
" \"vegetable\",\n",
" \"vegetable\",\n",
" \"vegetable\",\n",
" \"vegetable\",\n",
" ],\n",
" \"plts\": [\"fr\", \"fr\", \"fr\", \"fr\", \"ve\", \"ve\", \"ve\", \"ve\"],\n",
" \"quantity\": [\"1 kg\", \"10 kg\", \"1 kg\", \"10 kg\", \"1 kg\", \"10 kg\", \"1 kg\", \"10 kg\"],\n",
" \"product\": [\n",
" \"apple\",\n",
" \"apple\",\n",
" \"orange\",\n",
" \"orange\",\n",
" \"peppers\",\n",
" \"peppers\",\n",
" \"carrot\",\n",
" \"carrot\",\n",
" ],\n",
" \"price\": [1, 10, 2, 20, 1.5, 15, 1.5, 20],\n",
" \"price level\": [\"low\", \"low\", \"high\", \"high\", \"low\", \"low\", \"high\", \"high\"],\n",
" \"group\": [\n",
" \"fruit 1\",\n",
" \"fruit 10\",\n",
" \"fruit 1\",\n",
" \"veget\",\n",
" \"veget\",\n",
" \"veget\",\n",
" \"veget\",\n",
" \"veget\",\n",
" ],\n",
" \"id\": [1001, 1002, 1003, 1004, 1005, 1006, 1007, 1008],\n",
" \"supplier\": [\"sup1\", \"sup1\", \"sup1\", \"sup2\", \"sup2\", \"sup2\", \"sup2\", \"sup1\"],\n",
" \"location\": [\"fr\", \"gb\", \"es\", \"ch\", \"gb\", \"fr\", \"es\", \"ch\"],\n",
" \"valid\": [\"ok\", \"ok\", \"ok\", \"ok\", \"ok\", \"ok\", \"ok\", \"ok\"],\n",
"}"
]
},
{
Expand All @@ -57,7 +86,6 @@
"outputs": [],
"source": [
"import pandas as pd\n",
"import ntv_pandas as npd\n",
"\n",
"df_fruits = pd.DataFrame(fruits)\n",
"ana_fruits = df_fruits.npd.analysis(distr=True)"
Expand Down Expand Up @@ -92,9 +120,9 @@
}
],
"source": [
"print(ana_fruits.get_relation('plants', 'plts').typecoupl)\n",
"print(ana_fruits.get_relation('plants', 'product').typecoupl)\n",
"print(ana_fruits.get_relation('quantity', 'product').typecoupl)\n"
"print(ana_fruits.get_relation(\"plants\", \"plts\").typecoupl)\n",
"print(ana_fruits.get_relation(\"plants\", \"product\").typecoupl)\n",
"print(ana_fruits.get_relation(\"quantity\", \"product\").typecoupl)"
]
},
{
Expand Down Expand Up @@ -125,11 +153,11 @@
}
],
"source": [
"print('minimum distance: ', ana_fruits.get_relation('plants', 'plts').distance)\n",
"print('maximum distance: ', ana_fruits.get_relation('id', 'valid').distance)\n",
"print('intermediate distance: ', ana_fruits.get_relation('plants', 'product').distance)\n",
"# The 'plants' - 'product' relationship will be 'coupled' if we change, for example, \n",
"#'fruit-orange' in 'citrus-orange' and 'carrot-vegetable' in 'carrot-root vegetable' (2 changes) "
"print(\"minimum distance: \", ana_fruits.get_relation(\"plants\", \"plts\").distance)\n",
"print(\"maximum distance: \", ana_fruits.get_relation(\"id\", \"valid\").distance)\n",
"print(\"intermediate distance: \", ana_fruits.get_relation(\"plants\", \"product\").distance)\n",
"# The 'plants' - 'product' relationship will be 'coupled' if we change, for example,\n",
"#'fruit-orange' in 'citrus-orange' and 'carrot-vegetable' in 'carrot-root vegetable' (2 changes)"
]
},
{
Expand Down Expand Up @@ -164,7 +192,12 @@
],
"source": [
"# list of categories for each Field\n",
"print({field.idfield: category for field, category in zip(ana_fruits.fields, ana_fruits.category)})"
"print(\n",
" {\n",
" field.idfield: category\n",
" for field, category in zip(ana_fruits.fields, ana_fruits.category)\n",
" }\n",
")"
]
},
{
Expand Down Expand Up @@ -240,7 +273,7 @@
}
],
"source": [
"ana_fruits.partitions(mode='id')"
"ana_fruits.partitions(mode=\"id\")"
]
},
{
Expand Down Expand Up @@ -306,7 +339,7 @@
}
],
"source": [
"ana_fruits.field_partition() #first partition"
"ana_fruits.field_partition() # first partition"
]
},
{
Expand Down Expand Up @@ -364,7 +397,7 @@
}
],
"source": [
"ana_fruits.field_partition(partition=['product', 'quantity'])"
"ana_fruits.field_partition(partition=[\"product\", \"quantity\"])"
]
},
{
Expand Down Expand Up @@ -395,7 +428,7 @@
}
],
"source": [
"ana_fruits.relation_partition(partition=['product', 'quantity'])"
"ana_fruits.relation_partition(partition=[\"product\", \"quantity\"])"
]
},
{
Expand Down Expand Up @@ -433,7 +466,15 @@
"source": [
"from base64 import b64encode\n",
"from IPython.display import Image, display\n",
"display(Image(url=\"https://mermaid.ink/img/\" + b64encode(open('fruits.mmd', 'r', encoding=\"utf-8\").read().encode(\"ascii\")).decode(\"ascii\")))"
"\n",
"display(\n",
" Image(\n",
" url=\"https://mermaid.ink/img/\"\n",
" + b64encode(\n",
" open(\"fruits.mmd\", \"r\", encoding=\"utf-8\").read().encode(\"ascii\")\n",
" ).decode(\"ascii\")\n",
" )\n",
")"
]
},
{
Expand Down Expand Up @@ -871,7 +912,7 @@
"source": [
"from ntv_numpy import Xdataset\n",
"\n",
"xd_fruits = Xdataset.from_dataframe(df_fruits) # Xdataset is a neutral format\n",
"xd_fruits = Xdataset.from_dataframe(df_fruits) # Xdataset is a neutral format\n",
"xd_fruits.to_xarray(json_name=False)"
]
},
Expand All @@ -884,8 +925,12 @@
"source": [
"df2_fruits = xd_fruits.to_dataframe(json_name=False).reset_index()\n",
"\n",
"df_fruits_sort = df_fruits.sort_values(ana_fruits.partitions()[0]).reset_index(drop=True)\n",
"df2_fruits_sort = df2_fruits.sort_values(ana_fruits.partitions()[0]).reset_index(drop=True)[df_fruits.columns]"
"df_fruits_sort = df_fruits.sort_values(ana_fruits.partitions()[0]).reset_index(\n",
" drop=True\n",
")\n",
"df2_fruits_sort = df2_fruits.sort_values(ana_fruits.partitions()[0]).reset_index(\n",
" drop=True\n",
")[df_fruits.columns]"
]
},
{
Expand Down Expand Up @@ -1342,8 +1387,8 @@
}
],
"source": [
"xd_fruits_other = Xdataset.from_dataframe(df_fruits, dims=['product', 'quantity'])\n",
"xd_fruits_other.to_xarray(idxname=['product', 'quantity'])"
"xd_fruits_other = Xdataset.from_dataframe(df_fruits, dims=[\"product\", \"quantity\"])\n",
"xd_fruits_other.to_xarray(idxname=[\"product\", \"quantity\"])"
]
}
],
Expand Down

0 comments on commit 1e0bcfc

Please sign in to comment.