Combine Image Metadata

[mattmotoki]

Create a single file with a row for each of the following plants:

• Native Plants
• Invasive Plants
• Common Plants: hibiscus, plumeria, pineapple, ginger, anthurium, heliconia, bougainvillea

Add a column for each of the following fields:

• Hawaiian Name
• Common Name
• Genus
• Species
• Status (Endangered, Native, Endemic, Invasive)
• Story
• Uses

Upload all data to our shared data directory.

[mattmotoki]

Combined the Native, Invasive, Common plants into a single file called plant_categories.csv
The file is in our shared data directory.

All fields except for the "plant_name" (the category that we are predicting) need to be filled in.

[xyl012]

Input plant_categories_v2 with most fields filled in.

[mattmotoki]

Manually removed extra information; e.g., pineapple nutritional information. Cleaned up text see :

• replace new lines with space
• remove references [\d+]
• remove extra whitespace

For example,

The ovaries develop into berries, which coalesce into a large, compact, multiple fruit. The fruit of a pineapple is arranged in two interlocking helices, eight in one direction, 13 in the other, each being a Fibonacci number.[12]

The pineapple carries out CAM photosynthesis,[13] fixing carbon dioxide at night and storing it as the acid malate, then releasing it during the day aiding photosynthesis.

In the wild, pineapples are pollinated primarily by hummingbirds.[2][14] Certain wild pineapples are foraged and pollinated at night by bats.[15]

gets converted to

The ovaries develop into berries, which coalesce into a large, compact, multiple fruit. The fruit of a pineapple is arranged in two interlocking helices, eight in one direction, 13 in the other, each being a Fibonacci number. The pineapple carries out CAM photosynthesis, fixing carbon dioxide at night and storing it as the acid malate, then releasing it during the day aiding photosynthesis. In the wild, pineapples are pollinated primarily by hummingbirds. Certain wild pineapples are foraged and pollinated at night by bats. 

New data in plant_categories_v3.csv

[mattmotoki]

Added Sora's "canoe plants" to the list. These plants are pretty well-known in Hawaii and add about 500 images per plant to our dataset. The updated data is in local_meta_v4.csv.

I changed the name of the files because I am working with metadata from multiple sources. plant_categores and plant_meta was getting confusing. local will refer to the set of plants that will be predicted in our app.

Each plant in local_meta_v4.csv has a corresponding sub-folder in our shared images drive.