How to avoid NaN values when I use frame['Colum'].map(dict)

Question

I have the following dataset frame1

Color	Item
Red	Shirt
White	Shoes
Yellow	Shirt
Green	Shoes

I want to set all the colors for Shoes item to be "Blue", I use map

x = {"Shoes": "Blue"}
fr1["Color"] = fr1["Item"].map(x)

I expected the following result

Color	Item
Red	Shirt
Blue	Shoes
Yellow	Shirt
Blue	Shoes

Instead I got this

Color	Item
NaN	Shirt
Blue	Shoes
NaN	Shirt
Blue	Shoes

Answer 1

Either only use the rows with a valid key with boolean indexing:

x = {"Shoes": "Blue"}
m = fr1["Item"].isin(list(x))
fr1.loc[m, "Color"] = fr1.loc[m, "Item"].map(x)

Or fillna:

fr1["Color"] = fr1["Item"].map(x).fillna(fr1["Color"])