Complete the tasks below. Please turn in a single Jupyter notebook named 6_first_last.ipynb (substitute your first and last name). Please run Kernel > Restart & Run All on your notebook before turning in.

For this assignment, we will use Pandas to examine metadata from the Earth Microbiome Project.

First, download the metadata file for a 2,000-sample subset of the >27,000 samples in the Release 1 16S rRNA dataset.

curl -O "ftp://ftp.microbio.me/emp/release1/mapping_files/emp_qiime_mapping_subset_2k.tsv"

1. Import the tab-separated values file emp_qiime_mapping_subset_2k.tsv as a DataFrame called df with default data types, with the first row as column labels (columns) and the first column as row labels (indexes).
2. The indexes should be the sample IDs. How many samples are in this DataFrame? How many metadata columns?
3. What are the minimum and maximum pH values in the dataset?
4. What are the average and standard deviation temperature values in the dataset?

1. Make a new Series called temp with the temperature column as its own Series object. Remove NaN values (np.nan) from this Series. How many values are left?
2. Make a new DataFrame called df_seqs from columns sequences_split_libraries through observations_deblur_150bp (column positions 17-23) of the existing DataFrame. What is the mean value of column observations_deblur_90bp?
3. Save df_seqs as a csv file.

1. Store the first 5 rows of df_seqs as a new dataframe called df_seqs_head. Store the last 5 rows of df_seqs as a new dataframe called df_seqs_tail.
2. Concatenate df_seqs_head and df_seqs_tail using the concat() function.
3. Append df_seqs_tail to df_seqs_head using the append() function.
4. Make a new DataFrame called df_phys with the pH, temperature, and salinity columns (hint: you will need to know the exact column names; these are some of the last few columns). Make another new DataFrame called df_empo with the column empo_3 (note: this will actually be a Series because it has only one column, but you can treat it like a DataFrame).
5. Merge df_phys with df_seqs using the merge() function with the indexes of both DataFrames to make a new DataFrame called df_merged.
6. Join df_merged with df_empo using the join() function and store the result as df_merged.

1. Use a list comprehension to add a new column to df_merged called temperature_deg_f which takes the values in temperature_deg_c and converts them to degrees Fahrenheit.
2. Create a function that changes a single numerical value from Celsius to Fahrenheit. Apply this function to the values in temperature_deg_c using apply() to create a new column called temperature_deg_f_2.

1. Sort the rows in df_merged by sequences_split_libraries values from high to low and store the result as df_merged. (Hint: you can use inplace=True.)
2. Sort the columns in df_merged by column name from A to Z and store the result as df_merged. (Hint: you can use inplace=True.)